JP2012142488A - Semiconductor device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a semiconductor device in which an electrode of a semiconductor element is electrically connected to a connection pad via a solder bump and electromigration is effectively inhibited.SOLUTION: A semiconductor device comprises; a semiconductor element in which an electronic circuit 2 is directly connected to a part of a circumference of an electrode 3 of a semiconductor substrate 1; a wiring board 4 provided with a connection pad 5 arranged to face the electrode 3 of the semiconductor element; and a solder bump 6 sandwiched between the electrode 3 and the connection pad 5 and bonded to the electrode 3 and the connection pad 5. A nickel layer 7 lies along a boundary between the electrode 3 and the solder bump 6, and has a thickness larger than that at another portion, at a portion contacting a part of an outer periphery of the electrode 3, which is connected to the electronic circuit 2. A current density of a current flowing in the solder bump 6 is uniformed based on an electrical resistance difference depending on a difference in thicknesses of the nickel layer 7 and migration caused by current concentration can be inhibited.

Description

  The present invention relates to a semiconductor device in which a semiconductor element is mounted on a wiring board such as a ceramic wiring board, and electrodes on the main surface of the semiconductor element and connection pads on the main surface of the wiring board are joined to each other via solder bumps. Is.

  A semiconductor element such as a semiconductor integrated circuit element (IC) is usually mounted on a wiring board such as a ceramic wiring board for mounting a semiconductor element to form a semiconductor device, and an external electric circuit constituting a computer, communication equipment, sensor equipment, etc. (Motherboard etc.) is used by mounting.

  In general, a semiconductor element has a structure in which an electronic circuit is formed on one main surface of a semiconductor substrate such as silicon and a circular electrode or the like electrically connected to the electronic circuit is arranged on one main surface. The electronic circuit is electrically connected to the electrode by directly connecting one end thereof to a part of the outer periphery of the electrode.

  The wiring board has a structure in which connection pads are formed on the main surface of an insulating substrate such as a ceramic substrate so as to correspond to the arrangement of the electrodes of the semiconductor element. When the electrodes of the semiconductor element and the connection pads of the wiring substrate are opposed to each other and are joined to each other via solder bumps interposed therebetween, a semiconductor device is formed.

  The semiconductor device is mounted on the substrate of the electronic device, and various electric signals are transmitted and received between the electronic circuit of the semiconductor element and an external electric circuit. Along with the transmission and reception of the electronic signal, a current is passed between the electronic circuit of the semiconductor element and the solder bump.

  The electrodes of the semiconductor element are made of a metal material such as copper, aluminum, titanium, or nickel, and the solder bumps are made of solder such as tin-silver or tin-silver-copper.

Japanese Unexamined Patent Publication No. 7-211722 JP 2002-203925 A JP 2007-59937 A

  However, in such a conventional semiconductor device, the current (current density) as a signal transmitted and received between the electronic circuit and the solder bump is different in a part of the electrode where the electronic circuit is directly connected. It tends to be larger than this part. Therefore, in this part, there has been a problem that electromigration in which a metal material such as copper or aluminum forming the electrode diffuses and moves into the solder bump is likely to occur. When such electromigration occurs, voids are generated in a part of the electrode, causing problems such as local increase in electrical resistance and disconnection between the electrode and the solder bump.

  In particular, with the recent increase in the speed of semiconductor elements, the current flowing from the electronic circuit to the solder bumps tends to increase further. Therefore, the increase in current density in a part where the electronic circuit is connected is remarkable, and Migration is more likely to occur.

  The present invention has been completed in view of the above-mentioned problems of the prior art. In a semiconductor device in which electrodes of a semiconductor element are electrically connected to connection pads via solder bumps, the electrodes, particularly electronic circuits, An object of the present invention is to provide a semiconductor device capable of suppressing the generation of voids in which electromigration in the directly connected portion is effectively suppressed.

  A semiconductor device of the present invention includes a semiconductor element in which an electrode is disposed on a main surface of a semiconductor substrate on which an electronic circuit is formed, and the electronic circuit is directly connected to a part of the outer periphery of the electrode; A wiring board having a connection pad disposed opposite to the electrode; and a solder that is interposed between the electrode of the semiconductor element and the connection pad of the wiring board and bonded to the electrode and the connection pad A bump, and a nickel layer is interposed along the bonding interface between the electrode and the solder bump, the nickel layer being in contact with the part of the electrode to which the electronic circuit is connected It is characterized by being thicker than other parts.

  In the semiconductor device of the present invention having the above structure, the nickel layer contains phosphorus.

  In the semiconductor device according to the present invention, in any one of the configurations described above, the nickel layer is thickest at a portion in contact with the part to which the electronic circuit of the electrode is connected, and the distance from the part is It is characterized in that the farther away, the thinner the thickness.

  In the semiconductor device according to any one of the above-described structures, the nickel layer is thickest at a portion in contact with the part of the electrode and thicker than a central portion at an outer peripheral portion of the electrode. And

  According to the semiconductor device of the present invention, since the nickel layer is provided along the bonding interface between the electrode of the semiconductor element and the solder bump, the metal material that forms the electrode by the nickel layer has the above-described configuration. Diffusion to the solder bumps is suppressed. That is, the nickel layer functions as a so-called barrier layer against electromigration (hereinafter simply referred to as migration) of a metal material such as copper or aluminum, and suppresses the migration speed.

  In addition, this nickel layer is thicker at the part (connection part) in contact with the part where the electronic circuit of the electrode is connected than the other part, and has a higher electrical resistance (resistivity) than an electrode made of copper or the like. Therefore, the resistance of the electrical connection between the electrode and the solder bump at the connection portion (that is, the portion where the current density of the current flowing from the electronic circuit of the semiconductor element through the electrode to the solder bump tends to be the highest) But higher than the rest. Therefore, the difference in electrical resistance suppresses the current density from becoming larger at the connection portion than at other portions. Therefore, the void of the electrode in the said connection part is suppressed effectively. That is, the nickel layer functions as a barrier layer, and at the same time, adjusts the electrical resistance between the electrode and the solder bump to suppress the current density from becoming larger in one part of the electrode than in the other part. It also functions as a layer.

  In the semiconductor device of the present invention, in the above structure, when the nickel layer contains phosphorus, the electrical resistance (resistivity) of the nickel layer is larger than that of nickel alone due to the action of the phosphorus component. It has become. Therefore, the difference in electrical resistance caused by the difference in thickness of the nickel layer becomes larger, and the range of the current density that can be adjusted by the difference in thickness of the nickel layer (while keeping the difference in thickness of the nickel layer smaller) ) Can be larger. Therefore, in this case, the practicality of the nickel layer as the current density adjusting layer can be improved.

  In the semiconductor device of the present invention, in any of the above configurations, the nickel layer is thickest at a portion in contact with a part of the electrode, and the thickness is thinner as the distance from the part is farther, the electronic circuit The electrical resistance between the electrode and the solder bump can be increased in accordance with a decrease in the current density flowing from the solder to the solder bump. Therefore, it is possible to more effectively suppress the current density in a part of the electrode from being higher than that in the other part, thereby suppressing the generation of voids due to migration.

  The semiconductor device of the present invention has the above-described configuration when the nickel layer is thickest at the portion in contact with the part of the electrode and thicker than the central portion at the outer peripheral portion of the electrode. As in the case of, the current density at the connection portion is suppressed from becoming larger than that at other portions. Further, since the nickel layer is thick at the outer peripheral portion of the electrode, the central portion of the nickel layer is concave. Therefore, when solder bumps are joined to electrodes via a nickel layer therebetween, solder bumps (or solder balls that become solder bumps, etc.) are inserted by inserting the end portions of the solder bumps into the concave portions of the nickel layer. The positional deviation with respect to the electrode can also be suppressed.

It is sectional drawing which shows an example of embodiment of the semiconductor device of this invention. (A) is principal part sectional drawing which expands and shows the principal part of the semiconductor device shown in FIG. 1, (b) is the principal part bottom view (perspective view) which looked at the semiconductor element part of (a) from the lower surface side. Figure). (A) And (b) is principal part sectional drawing which expands and shows the principal part in the other example of embodiment of the semiconductor device of this invention, respectively. (A) And (b) is a principal part bottom view (perspective view) which expands and shows the principal part in the other example of embodiment of the semiconductor device of this invention, respectively. It is the disassembled perspective view which looked at the semiconductor element part of the semiconductor device shown to Fig.4 (a) from the lower surface side.

  A semiconductor device of the present invention will be described with reference to the accompanying drawings. FIG. 1 is a sectional view showing an example of an embodiment of a semiconductor device of the present invention. FIG. 2A is an enlarged cross-sectional view showing an essential part of the semiconductor device shown in FIG. 1, and FIG. 2B is a view of the semiconductor element portion of FIG. It is the principal part bottom view (perspective view). 1 and 2, 1 is a semiconductor substrate, 2 is an electronic circuit, 3 is an electrode, 4 is a wiring board, 5 is a connection pad, 6 is a solder bump, and 7 is a nickel layer. The semiconductor device is basically configured by electrically and mechanically connecting the semiconductor substrate 1 including the electronic circuit 2 and the electrodes 3 and the wiring substrate 4 including the connection pads 5 via the solder bumps 6. In FIG. 1, the nickel layer 7 is omitted. FIG. 2B shows a state in which the wiring board 4 and the solder bump 6 are omitted and the electrode 3 portion of the semiconductor element is seen through from the lower surface side.

  The semiconductor substrate 1 is made of a semiconductor material such as silicon, phosphorus gallium arsenide, germanium, gallium arsenide, gallium nitride, or silicon carbide, and is manufactured by cutting a wafer of such a semiconductor material into a predetermined shape and size. . The semiconductor substrate 1 is, for example, a quadrangular plate-like silicon substrate having a side length of about 3 to 10 mm.

  An electronic circuit 2 made of a metal material such as aluminum, copper, or gold is disposed on the main surface of the semiconductor substrate 1. The electronic circuit 2 includes a part having functions such as a transistor and a diode, for example, and is formed as an integrated circuit on the main surface of the semiconductor substrate 1.

The electronic circuit 2 is directly connected to a part of the outer periphery of the electrode 3 and is electrically connected to an external electric circuit via the electrode 3. The semiconductor substrate 1, the electronic circuit 2, and the electrode 3 form a semiconductor element (no symbol) such as a semiconductor integrated circuit element (IC). The electrode 3 is made of a metal material such as copper, aluminum, silver, palladium, or nickel. Of these metal materials, copper and aluminum are often used as the main components of the metal material forming the electrode 3 in consideration of low electrical resistance, economy and the like.

  The shape and dimensions of the electrode 3 are appropriately set according to the arrangement position of the electronic circuit 2 and the like, for example, are formed in a circular shape having a diameter of about 100 to 300 μm.

  For example, after forming a silicon oxide film (not shown) on the main surface of a semiconductor substrate 1 (actually a silicon wafer or the like), a semiconductor element is formed by applying an electronic circuit 2 made of aluminum or the like by vapor deposition or photolithography. It is manufactured by applying a fine processing technique.

  Signals are transmitted and received between the electronic circuit 2 and an external electric circuit via the electrode 3, and various processes such as predetermined computation, storage, and analysis are performed in the electronic circuit 2.

  The semiconductor element is electrically and mechanically connected to the wiring board 4 as described above, that is, mounted on the wiring board 4 to become a semiconductor device, and is used as a component in various electronic devices such as computers, communication devices, and inspection devices. Implemented. A circuit such as a motherboard provided in the electronic device corresponds to an external electric circuit.

  The wiring substrate 4 is a so-called IC package or the like for mounting a semiconductor element to manufacture a semiconductor device. The wiring substrate 4 is made of, for example, a ceramic material such as an aluminum oxide sintered body, a glass ceramic sintered body, a mullite sintered body, an aluminum nitride sintered body, a resin material such as an epoxy resin or a polyimide resin, or a ceramic material. In addition, connection pads 5 are formed on the main surface of an insulating substrate 4a formed of an insulating material such as a composite material of a glass material or the like and a resin material.

When the insulating substrate 4a is made of, for example, a glass ceramic sintered body, it can be manufactured as follows. In other words, a suitable organic binder and organic solvent are added to and mixed with the raw material powder obtained by adding glass and sintering aids to aluminum oxide to form a slurry, and this is applied to a sheet molding technique such as a doctor blade method or a lip coater method. A plurality of ceramic green sheets are obtained by forming into a sheet shape, and then the ceramic green sheets are formed into an appropriate shape by cutting or punching, and a plurality of them are laminated, and finally the laminated ceramic green sheets The sheet is manufactured by firing at a temperature of about 800 to 1000 ° C. in a reducing atmosphere.

  The connection pad 5 is disposed on the main surface (upper surface in the example of FIG. 1) of the insulating substrate 4a so as to face the electrode 3 disposed on the main surface (lower surface in the example of FIG. 1). By connecting the connection pad 5 and the electrode 3 of the semiconductor element to each other via the solder bump 6, the semiconductor element and the wiring board 4 are electrically and mechanically connected to form a semiconductor device.

  The connection pad 5 is electrically connected to an external electric circuit (no symbol) through a wiring conductor 8 formed on the surface such as the lower surface of the insulating substrate 4a or inside, for example. The wiring conductor 8 includes a through conductor (via conductor) formed inside the insulating substrate 4a. The electrical connection between the wiring conductor 8 and an external electric circuit is performed through a conductive connecting material 9 such as solder or a conductive adhesive. The wiring conductor 8 can be formed by the same method using, for example, the same material as the connection pad 5 described later.

  The connection pad 5 is made of a metal material such as tungsten, molybdenum, manganese, copper, silver, palladium, gold, platinum, or titanium. Since these metal materials, particularly copper and silver, have low electrical resistance, electrical characteristics such as conduction resistance when the electronic component 11 is electrically connected to the electrode pad 2 by suppressing the electrical resistance in the electrode pad 2 low. It is advantageous to make it high.

  The connection pad 5 has, for example, a circular shape, an elliptical shape, a quadrangular shape, or the like in plan view, and the shape and size are appropriately set according to the shape and size of the electrodes 3 connected to face each other. For example, if the electrode 3 has a circular shape with a diameter of about 100 to 300 μm as described above, the connection pad 5 is formed in a slightly larger circular shape (with a diameter of about 200 to 400 μm).

  The solder bumps 6 that connect the electrodes 3 of the semiconductor element and the connection pads 5 of the wiring board 4 are flattened vertically (for example, on the electrode 3 side and the connection pad 5 side) as shown in FIGS. It is spherical.

  The solder bump 6 is formed of a solder material such as tin-lead, tin-silver, tin-silver-copper, tin-silver-bismuth or the like. The solder bump 6 is preferably a so-called lead-free solder that does not contain harmful substances such as lead in order to avoid adverse effects on the environment.

  The connection between the electrode 3 and the connection pad 5 via the solder bump 6 is performed as follows, for example. That is, first, a solder ball such as tin-silver solder is prepared, and this solder ball is positioned and placed on the electrode 3 (with the semiconductor element set so that the electrode 3 faces upward). It is heated by a method such as reflow and bonded to the electrode 3, and then the semiconductor element is turned over and mounted on the wiring board 4 so that the solder bumps 6 are placed on the connection pads 5. By heating the bump 6 again and bonding it to the connection pad 5, the electrode 3 and the connection pad 5 are connected to each other through the solder bump 6. And the semiconductor device by which the semiconductor element and the wiring board 4 are electrically and mechanically connected through the solder bump 6 by the above process is manufactured.

  A nickel layer 7 is interposed between the solder bump 6 and the electrode 3 along the interface of these joints. In addition, the nickel layer 7 is thicker than the other portions in a portion in contact with a part to which the electronic circuit 2 on the outer periphery of the electrode 3 is connected.

  Since the nickel layer 7 is interposed along the bonding interface between the electrode 3 of the semiconductor element and the solder bump 6, the metal material forming the electrode 3 by the nickel layer 7 (particularly, copper or the main component) Diffusion of aluminum) into the solder bumps 6 is suppressed. That is, the nickel layer 7 functions as a so-called barrier layer against migration of metal materials such as copper and aluminum (as described above, electromigration which is diffusion of metal components due to electron movement), and the migration speed. Suppress.

  Further, the nickel layer 7 has a higher electrical resistance (resistivity) than the electrode 3 made of copper or the like, and is in a portion in contact with a part (connection portion) of the electrode 3 where the electronic circuit 2 is directly connected. Since it is thicker than the other parts, the electrode 3 and the solder bump in the connection part (that is, the part where the current density of the current flowing from the electronic circuit 2 of the semiconductor element through the electrode 3 to the solder bump 6 tends to be the highest). The resistance of the electrical connection with 6 is higher than the other parts. Therefore, the difference in electrical resistance suppresses the current density from becoming larger at the connection portion than at other portions. Therefore, the void of the electrode 3 in the connection portion is effectively suppressed.

That is, the nickel layer 7 functions as a barrier layer that suppresses the diffusion of the components of the electrode 3 such as copper, and at the same time, the electric density between the electrode 3 and the solder bump 6 is adjusted so that the current density is a part of the electrode 3. It also functions as a layer for adjusting the current density, which is suppressed from becoming larger than other portions.

  Nickel has good adhesion to any of solder materials such as copper, aluminum, and tin-silver solder. Therefore, the nickel layer 7 has good adhesion to both the electrode 3 and the solder bump 6, and the mechanical bonding strength between them is also high. Therefore, in the configuration in which the nickel layer 7 is interposed between the electrode 3 and the solder bump 6, it is possible to ensure high reliability of electrical and mechanical connection between the electrode 3 and the solder bump 6.

  In order to interpose the nickel layer 7 between the electrode 3 and the solder bump 6, for example, when a semiconductor device is manufactured by the above-described process, before the solder ball is placed on the electrode 3, the nickel layer is formed on the surface of the electrode 3. 7 may be deposited by a method such as electroless plating, and then the solder ball may be bonded to the surface of the electrode 3 (actually the nickel layer 7) by the same method as described above.

  The nickel layer 7 is formed of, for example, nickel or a material containing nickel as a main component and components such as other metals. Other components such as metal are components that are eutectoid when the nickel layer 7 is deposited on the electrode 3 by a method such as electroless plating as described above. When the nickel layer 7 is formed on the surface of copper or the like, which is the material of the electrode 3, the electroless nickel is used in the case where importance is attached to the adhesion with copper, the formation cost, the chemical resistance, and the mountability. -Phosphorus plating is used.

  When the nickel layer 7 is formed by an electroless plating method (electroless nickel-phosphorus), the electrode 3 of the semiconductor substrate 1 is mainly composed of a nickel salt such as nickel sulfate and a reducing agent such as sodium hypophosphite, It can be formed by immersing in an electroless nickel plating solution to which a complexing agent, a stabilizer and the like are added for a predetermined time. In this case, the formed nickel layer 7 contains about 5 to 10% by mass of phosphorus which is a decomposition product of the reducing agent.

The electrical resistance (μΩ · cm) of these metals is approximately 6.9 for nickel, approximately 1.694 for copper, and approximately 2.67 for aluminum, and the electrical resistance of nickel is higher than that of copper and aluminum. The resistance can be increased. When the nickel layer 7 is formed by electroless plating, the nickel plating film as the nickel layer 7 is a nickel-phosphorus alloy containing phosphorus. The nickel-phosphorus alloy has an electrical resistance of 50-60 μΩ · cm at a phosphorus concentration of about 8% by mass, which is higher than the solder material tin of about 12.6 μΩ · cm. The resistance value can be increased.

  In other words, the nickel layer 7 preferably contains phosphorus in consideration of making the function as a current density adjusting layer between the electrode 3 and the solder bump 6 more effective. The nickel layer 7 containing phosphorus can be formed, for example, by depositing the electrode 3 with a nickel plating layer that becomes the nickel layer 7 by electroless nickel-phosphorus plating as described above.

  The phosphorus content in the nickel layer 7 is preferably as much as possible in order to increase the electrical resistance of the nickel layer 7. However, if the phosphorus content is too high, there is a possibility that problems such as a decrease in mounting reliability (connection reliability between the solder bump 6 and the nickel layer 7) during solder mounting may be induced. Accordingly, the phosphorus content in the nickel layer 7 is preferably in the range of 5 to 10% by mass.

  In order to adjust the phosphorus content within the above range for the nickel layer 7, it is important to control the pH, for example, using the above electroless nickel-phosphorous plating solution. Since the phosphorus content decreases as the pH increases, the pH may be controlled to about 4.4 to 4.8 for an acidic type nickel plating solution and about 6.0 to 6.8 for a neutral type.

  In addition, when making the nickel layer 7 contain phosphorus, it is preferable that the distribution of phosphorus in the nickel layer 7 is substantially uniform.

  Examples of the material added to the nickel layer 7 in order to increase the electrical resistance (resistivity) in the nickel layer 7 include a method of adding a resin such as PTFE (polytetrafluoroethylene) in addition to phosphorus. In the method of adding these phosphorus and resin to the nickel layer 7, for example, by adding and dispersing resin or the like in the form of particles in the plating solution, plating can be performed in the nickel layer 7.

  The thickness of the nickel layer 7 needs to be thicker than the other portions in the portion in contact with the connection portion of the electrode 3, but the specific thickness in each portion is from the electronic circuit 2 through the electrode 3 to the solder bump 6. Conditions affecting the migration, such as the magnitude of the current flowing through the substrate, the temperature of the environment in which the semiconductor device is used, the electrical resistance (resistivity) of the nickel layer 7 itself, the composition of the metal material forming the electrode 3 And may be set as appropriate according to productivity, economic efficiency, and the like.

For example, the current density flowing from the electronic circuit 2 through the electrode 3 is 8 kA / cm per electrode 3.
² and the electrode 3 is made of copper and the nickel layer 7 contains about 5% by mass of phosphorus, the thickness of the nickel layer 7 is about 8 to 15 μm at the portion in contact with the connecting portion of the electrode 3. In other parts, it may be about 2 to 8 μm.

  In this case, by providing the nickel layer 7 with a thick portion as described above, in other words, according to the ease of current flow, a portion having a high resistance is provided between the electrode 3 and the solder bump 6. By providing, the current density of the current flowing from the electrode 3 to the solder bump 6 can be made substantially the same in the entire area of the electrode 3.

  As a simple method, there is a method in which the thickness of the other portion of the nickel layer 7 is set to about 4 μm, and a thickness of about 2 to 4 times the thickness is set as the thickness of the portion in contact with the connection portion of the electrode 3.

  For the nickel layer 7, for example, a masking method can be used to form a thicker portion than the other portions as described above.

  That is, first, the entire surface of the electrode 3 is made relatively thin (for example, to a thickness of about 4 μm in other portions as described above) by a method such as electroless plating (first layer). Thereafter, the surface of the first layer is covered with a masking material except for a portion in contact with the connection portion of the electrode 3 (that is, a portion where the nickel layer 7 is thickened). As the masking material, for example, a novolak resin or the like used for resist processing can be used, and it can be applied to a predetermined pattern by a photolithography method. After that, nickel plating is added to the surface of the first layer exposed without being covered with this masking material, and if the masking material is removed after setting the thickness to a predetermined thickness, other portions are in contact with the connection portion of the electrode 3. It is possible to form a nickel layer 7 thicker than this portion.

  The portion where the thickness of the nickel layer 7 is increased always includes the portion in contact with the connection portion of the electrode 3, and from this portion, for example, the difference in the thickness of the nickel layer 7 due to the magnitude of current density or masking as described above. What is necessary is just to set suitably according to workability | operativity at the time of attaching. In the example shown in FIG. 2, the nickel layer 7 is thicker than the other portions in a circular arc-shaped portion having a constant width along the outer periphery of the electrode 3 including the connection portion in plan view.

Further, for example, as shown in FIGS. 3A and 3B, this semiconductor device is thickest at a portion where the nickel layer 7 is in contact with a part of the electrode 3 and the distance from the part is farther. Is gradually reduced, the electrical resistance between the electrode 3 and the solder bump 6 can be gradually increased in accordance with the decrease in the current density flowing from the electronic circuit 2 through the electrode 3 to the solder bump 6. Therefore, it is possible to more effectively suppress the current density in part of the electrode 3 from becoming larger than that in other parts, and to suppress the generation of voids due to migration. 3 (a) and 3 (b) are enlarged cross-sectional views showing a main part in another example of the embodiment of the semiconductor device of the present invention. In FIG. 3, the same parts as those in FIGS. 1 and 2 are denoted by the same reference numerals.

  In other words, in this case, the electric resistance between the electrode 3 and the solder bump 6 is gradually increased in accordance with a current that is the largest at the connection portion of the electrode 3 and tends to gradually decrease as the distance from the connection portion increases. As a result, the current density can be more reliably made equal to the same level in the entire area of the electrode 3. As a result, gaps in the electrode 3 due to migration caused by an excessive current density can be suppressed.

  Such an effect is the same when the nickel layer 7 contains phosphorus. When the nickel layer 7 contains phosphorus, the rate of change in electrical resistance according to the change in the thickness of the nickel layer 7 becomes larger. Further, the thickness of the nickel layer 7 necessary for obtaining the same electrical resistance can be further reduced.

  FIG. 3A shows an example in which the thickness of the nickel layer 7 is gradually reduced from the portion in contact with the connection portion of the electrode 3 to the other portion. In the case of this example, the electrical resistance in the nickel layer 7 can be adjusted with higher accuracy in accordance with the tendency of current increase and decrease. Therefore, it is advantageous in adjusting the current density with higher accuracy.

  FIG. 3B shows an example in which the thickness of the nickel layer 7 is changed stepwise. In this case, the nickel layer 7 can be formed by a method such as an electroless plating method by a method using masking in combination with each thickness. Therefore, in this case, as described above, it is advantageous in improving workability when the thickness of the nickel layer 7 is gradually changed from the connection portion of the electrode 3 to another portion.

  Further, this semiconductor device may have a form as shown in FIGS. 4A and 4B, for example. FIGS. 4A and 4B are enlarged bottom views (perspective views) showing main parts in another example of the embodiment of the semiconductor device of the present invention. 4, parts similar to those in FIGS. 1 and 2 are denoted by the same reference numerals. The effects obtained in the examples shown in FIGS. 4A and 4B can be obtained similarly when the nickel layer 7 contains phosphorus.

  For example, as shown in FIG. 4A, the semiconductor device has the above thickness when the nickel layer 7 is thickest at the portion in contact with the connection portion of the electrode 3 and thicker at the outer peripheral portion of the electrode 3 than the central portion. As in the case of the configuration, the current density at the connection portion is suppressed from being larger than that at other portions. In this case, since the nickel layer 7 is thick at the outer periphery of the electrode 3, the central portion of the nickel layer 7 is concave. Therefore, when the solder bump 6 is joined to the electrode 3 via the nickel layer 7, the end portion of the solder bump 6 is inserted into the concave portion of the nickel layer 7, and the position of the solder bump 6 with respect to the electrode 3. The deviation can also be suppressed.

  That is, as shown in FIG. 7, the solder bumps 6 were heated by positioning the solder balls 6a such as tin-silver solder to be the solder bumps 6 on the electrodes 3 (nickel layer 7). Is. On the other hand, in the case of the example shown in FIG. 7A, a part of the lower side of the solder ball 6a is fitted to the concave portion in the center of the nickel layer 7 and temporarily fixed so as not to roll. can do. Therefore, it is advantageous in improving productivity as a semiconductor device.

  In addition, as shown in FIG. 4B, for example, the semiconductor device 7 has a nickel layer 7 that is thicker than other portions in a fan-shaped range centering on a portion in contact with the connection portion of the electrode 3. The electric resistance of the nickel layer 7 can be gradually increased in accordance with the current that tends to be highest at this center and tends to decrease to the same value at the same distance from the center. Therefore, in this case, it is possible to obtain a semiconductor device suitable for effectively suppressing the current density in part of the electrode 3 from being larger than that in other parts.

  The following semiconductor element and wiring board are prepared, and the electrodes of the semiconductor element and the connection pads of the wiring board are electrically and mechanically connected to each other via solder bumps made of tin-silver solder, and the semiconductor device of the embodiment And the semiconductor device of the comparative example was produced. In the semiconductor device of the example, the following nickel layer was interposed between the solder bump and the electrode, and the nickel layer was not interposed in the semiconductor device of the comparative example (the electrode and the solder bump were directly connected). .

Semiconductor element: As a semiconductor substrate, a silicon substrate having a square plate with a side length of 5 × 5 mm is used, and an electronic circuit made of aluminum and an electrode made of copper are formed on the main surface of the semiconductor substrate through a silicon oxide film. The arranged one was used. The number of electrodes was 64, and the electrodes were arranged vertically and horizontally on the main surface of the semiconductor substrate. The electrode was circular with a diameter of about 100 μm, and the current flowing through the electrode was 8 kA / cm ² .

Wiring substrate: The diameter of the main surface (upper surface) of a square plate-shaped insulating substrate made of a glass ceramic sintered body with a side length of about 10 x 10 x 1 mm is formed by a tungsten (copper) metallization layer. Formed a circular connection pad of about 150 μm. The number of connection pads is 64, the same as the number of electrodes of the semiconductor element, and each connection pad was formed at a position corresponding to the electrode of the semiconductor element.

Solder bump: Tin-silver-copper (Sn-3Ag-0.5Cu) solder was used. The solder bumps were formed by placing solder balls having the above composition on the electrodes of the semiconductor element and heating and melting them at about 260 ° C.

Nickel layer: In the semiconductor device of the example, before joining the solder bump to the electrode, the surface of the electrode was nickel using an electroless nickel plating solution with sodium hypophosphite as a reducing agent and adjusted to pH 4.6. A layer was formed. The phosphorus content of the nickel layer was about 5% by mass. The thickness of the nickel layer was about 16 μm at the part (thick part) in contact with the electrode connection part, and about 4 μm at the other part. The difference in thickness of the nickel layer was performed by a method using a masking material. That is, a nickel layer having a thickness of 4 μm is first deposited on the entire surface of the electrode, and then a portion other than the portion in contact with the electrode connection portion is masked by photolithography using a novolac resin, so that the thickness of the thick portion becomes 16 μm. Thus, nickel plating was performed additionally. In this embodiment, the range of the portion where the nickel layer is thick and in contact with the connection portion of the electrode is a fan-shaped range having a radius of about 50 μm with the connection portion as the center.

After mounting the semiconductor device and the circuit of the printed circuit board (external electric circuit) for 500 hours after mounting each of the semiconductor devices in the above examples and comparative examples on the printed circuit board. Calculate the rate of increase in resistance before and after energization of the electrode, and increase rate 20%
With the above, it was determined that there was a failure.

  As a result, no failure occurred in the semiconductor device of the example of the present invention, whereas failure occurred in 20% of the electrodes in the semiconductor device of the comparative example. Thereby, the effect which suppresses generation | occurrence | production of the space | gap in an electrode in the semiconductor device of this invention has been confirmed.

DESCRIPTION OF SYMBOLS 1 ... Semiconductor substrate 2 ... Electronic circuit 3 ... Electrode 4 ... Wiring board 4a ... Insulation board 5 ... Connection pad 6 ... Solder bump 6a ... Solder ball 7 ... Nickel Layer 8 ... Wiring conductor 9 ... Conductive connecting material

Claims (4)

An electrode is disposed on a main surface of a semiconductor substrate on which an electronic circuit is formed, and a semiconductor element in which the electronic circuit is directly connected to a part of the outer periphery of the electrode, and the semiconductor element is disposed to face the electrode of the semiconductor element A wiring board provided with a connection pad, and a solder bump interposed between the electrode of the semiconductor element and the connection pad of the wiring board and joined to the connection pad. A nickel layer is interposed along the bonding interface between the solder bump and the nickel bump, and the nickel layer is thicker than the other part in the part of the electrode in contact with the part to which the electronic circuit is connected. A semiconductor device characterized by the above. The semiconductor device according to claim 1, wherein the nickel layer contains phosphorus. The nickel layer is thickest at a portion of the electrode that is in contact with the part to which the electronic circuit is connected, and the thickness of the nickel layer decreases as the distance from the part increases. 2. The semiconductor device according to 2. 3. The semiconductor device according to claim 1, wherein the nickel layer is thickest at a portion in contact with the part of the electrode, and is thicker than a central portion at an outer peripheral portion of the electrode.

Images