Copper bonding wires and semiconductor devices using the same, metal oxide semiconductor field-effect transistors, inverters, insulated gate bipolar transistors, lithium-ion battery modules, electric vehicles

Abstract

The present invention provides a copper bonding wire that improves cutting ability, bonding strength, and conformability, while suppressing wire lift-off and tool detachment. [Solution] A copper wire made of a copper alloy with a copper purity of 99.99% by mass or higher, wherein the grain boundary density in a cross-section perpendicular to the wire axis of the copper wire is 0.01 μm -1 Above 0.6 μm -1 Less than and among the crystal orientations in the wire axis direction in the cross-section, the crystal orientation having an angular difference of 15° or less with respect to the wire axis direction. <111> The orientation ratio of the crystal orientation <101> A copper bonding wire having a value obtained by dividing by the orientation ratio of 10 or more and 650 or less, a dynamic hardness of 45 or more and 90 or less, and an elastic modulus of 20 GPa or more and 70 GPa or less in the cross-section.

Description

[Technical Field]

[0001] The present invention relates to copper bonding wires (hereinafter referred to as "copper wires"), and more particularly to copper bonding wires suitable for applications involving high currents, such as power semiconductors and connecting busbars and electrodes of lithium-ion batteries in lithium-ion battery mobile devices for electric vehicles, as well as a method for manufacturing the same and a semiconductor device using the copper wire.

[0002] Generally, "semiconductors" refer to a group of components such as CPUs (Central Processing Units) and memory, whose main roles are "computation" and "memory." Semiconductors are used in consumer electronics such as personal computers, smartphones, and televisions. Power semiconductors, on the other hand, are responsible for driving motors, charging batteries, and supplying power to operate microcomputers and LSIs (Large-Scale Integrated Circuits). Power semiconductors are mainly used for changing voltage and frequency, and for power conversion (converting DC to AC or AC to DC). Power semiconductors are also called power semiconductors, power devices, power elements, or power semiconductor elements.

[0003] Inverters, found in energy-saving home appliances such as air conditioners, refrigerators, and washing machines, are a familiar example of the use of power semiconductors. Inverters control the rotation speed of a motor by converting its frequency. By freely changing the rotation speed of the motor, inverters reduce unnecessary motor movement and contribute to energy saving. On the other hand, air conditioners without inverters adjust the room temperature by repeatedly starting and stopping the motor, which can lead to problems such as unstable temperature control and high power consumption. These functions of inverters are achieved by power semiconductors (power transistors) performing "switching," which precisely switches the current on and off.

[0004] Power semiconductors are widely used not only in energy-saving home appliances but also in the transportation sector, such as electric vehicles and hybrid vehicles. In the transportation sector, power semiconductors such as IGBTs (Insulated Gate Bipolar Transistors) are used to perform power conversion and control. An IGBT consists of bonding wires that connect IGBT chips (power chips) to each other and to external electrodes. Bonding wires for power semiconductors are made of relatively thick aluminum wires with a diameter of 40 μm to 700 μm because large currents flow through them.

[0005] As described above, general semiconductors, whose primary purpose is low-power calculations, and power semiconductors, whose primary purpose is to control power itself using high power, are considered to be completely different, even though they are both semiconductors. Bonding wires used in general semiconductors only carry small currents. Therefore, bonding wires used in general semiconductors are often thin wires made of materials such as Au, Ag, or Cu, with a wire diameter of 30 μm or less. The bonding method involves melting one end of the wire to form a molten ball, and then applying heat, ultrasound, and pressure to bond it to the semiconductor chip electrode, etc. In contrast, bonding wires used in power semiconductors need to carry large currents, so as mentioned above, relatively inexpensive aluminum wires with a wire diameter of 40 μm or more and low electrical resistance are used. In the bonding method for bonding wires used in power semiconductors, a molten ball is not created, and the wire is bonded to the electrode, etc., by applying ultrasound and pressure to the wire at room temperature.

[0006] Furthermore, unlike general semiconductors, power semiconductors, as mentioned above, repeatedly switch on and off high currents, requiring bonding wires that can withstand thermal loads from heat generation and cooling. In addition, because they are used in applications where extremely high safety standards are required, such as automobiles and trains, high levels of durability and reliability are essential. For these reasons, bonding wires used in general semiconductors and bonding wires used in power semiconductors have completely different required characteristics and properties, even though they are both bonding wires.

[0007] In recent years, as a measure against global warming and climate change, an international framework (the Paris Agreement) was adopted at the COP21 international conference on reducing greenhouse gas emissions, including the reduction of greenhouse gas emissions from 2020 onwards, accelerating the global trend towards a decarbonized society. The automotive industry plays a major role in this. European countries have announced a policy to completely ban the sale of internal combustion engine vehicles (gasoline and diesel cars) by 2035 and limit sales to electric vehicles (including plug-in hybrid vehicles). The United States has also set a target of having electric vehicles account for 50% of its new car market by 2030. This trend is similar in emerging countries such as China and India, and Japan also plans to switch all new car sales to electric vehicles (including hybrid vehicles) by 2035.

[0008] As the shift to electric vehicles accelerates, there are challenges hindering their widespread adoption. These include short driving range (the distance a vehicle can travel from a full charge until the battery runs out) and a limited number of charging facilities (locations).

[0009] In particular, improving the performance of power semiconductors is essential for extending the driving range. Electric vehicles can run by converting the direct current from the battery, which is the energy source, into alternating current using an inverter, which is a power semiconductor, to drive the motor.

[0010] Until recently, power semiconductors were made of silicon (Si), but in recent years, power semiconductors made of silicon carbide (SiC) have been released. Conventional Si power semiconductors experience energy loss when power is applied and when switching on and off. All of this electrical energy loss is released as heat (heat generation) from the inverter. A heat dissipation mechanism is required to suppress this heat generation, which has resulted in a reduction in the passenger space of electric vehicles, an increase in vehicle weight, and a reduction in driving range.

[0011] The application of SiC power semiconductors to electric vehicles enables a reduction in energy loss both during energization and during the on / off switching operation. As a result, passive elements such as transformers, which benefit from increased switching frequency (higher frequency), can be miniaturized, and power semiconductors such as inverters can be made smaller and lighter. In addition, the simplification of the heat dissipation mechanism leads to a more compact (lighter) electric vehicle as a whole, and the high-efficiency operation of the inverter extends the driving range on a single charge.

[0012] Furthermore, while the widespread adoption of battery charging facilities is essential for the proliferation of electric vehicles, there is also a growing need for rapid charging. As a result, bonding wires that can withstand the high currents generated by rapid charging are required.

[0013] Amidst the trend toward miniaturization and high-current handling, most bonding wires in power semiconductors currently use aluminum-based materials. Aluminum bonding wires are used not only in power semiconductors but also as wires connecting the busbars of lithium-ion battery mobile devices in electric vehicles to the electrodes of lithium-ion batteries. Because aluminum has high thermal and electrical resistance, it generates a lot of heat when current is passed through it, and its heat dissipation effect is low. Therefore, the adoption of aluminum wires goes against the trend toward miniaturization of power semiconductors. Furthermore, it has been reported that in the weak bonding areas of aluminum wires, grain coarsening occurs due to the heat generated when high currents are passed through, causing the bonding area to break.

[0014] Therefore, copper wire is attracting attention as an alternative material to aluminum wire. Compared to aluminum wire, copper wire has lower thermal and electrical resistance, thus suppressing heat generation, and its higher recrystallization temperature allows it to withstand high currents.

[0015] Meanwhile, not only in the electric vehicle industry, but also in AC adapters used to charge smartphones and laptops, semiconductors made of gallium nitride (GaN), known as wide-bandgap semiconductors, are increasingly being incorporated. By using wide-bandgap semiconductors, it is possible to downsize them to about half the size of conventional designs, achieving ultra-miniaturization.

[0016] Furthermore, power semiconductors are playing an increasingly important role in the Internet of Things (IoT), a hot topic these days. As IoT-equipped home appliances become smaller, thinner, and more densely packed year after year, power semiconductors are also becoming smaller, thinner, and more densely packed, resulting in less space for bonding the power chip and bonding wires. Therefore, bonding wires must be joined in a way that makes the most of the limited space available.

[0017] As mentioned earlier, copper wire generates less heat than aluminum wire and can withstand high power, but its hardness and higher modulus of elasticity compared to aluminum are major obstacles to the miniaturization of power semiconductors.

[0018] For example, the bonding of the bonding wire to the power chip (first bond) and the bonding wire to the external electrode (second bond) are generally performed in the same direction and in an almost straight line. However, with the miniaturization of power semiconductors, the second bond must be made targeting the small gap around the electrode, which necessitates sharp bending and twisting, and bonding at a sharp angle. Therefore, the bonding wire needs a degree of flexibility in bending, allowing the wire direction (angle) to be changed during the bonding operation. In other words, flexible conformability is a very important characteristic for bonding wires when miniaturizing power semiconductors. The same challenge exists in so-called reverse wire bonding, where the external electrode side is the first bond and the power chip side is the second bond. In this regard, the elastic modulus of copper wire is higher than that of aluminum wire, so the conformability of copper wire is very poor. Therefore, there is a higher possibility of the copper wire coming off the wedge tool than with aluminum wire.

[0019] As described above, the inventors have discovered that when the wire connecting the first joint and the second joint is forced to twist or bend sharply, it also has an adverse effect on the cutting property for cutting the wire. That is, due to sharp bending or the like, the wire is slightly deformed, so that the cutter cannot properly enter the wire and the wire cannot be completely cut, so-called cut residue may occur. When cut residue occurs, peeling (lift-off) occurs from the joint before the wire is torn off, or although lift-off does not occur, cutting cannot be performed below the height set for the bonder, the height limit works, and a problem that the device stops occurs.

[0020] To summarize the above, in order to reduce greenhouse gas emissions, a shift to electric vehicles is essential, and it is urgent to (A) extend the cruising range and (B) increase the production of rapid charging facilities, which are the keys to the popularization of these electric vehicles. Regarding (A), weight reduction of the vehicle is necessary, and for this purpose, miniaturization of the power semiconductor and high heat dissipation are essential, so the need for copper bonding wires with low electrical resistance and low thermal resistance is increasing. Regarding (B) as well, it is essential to adopt copper bonding wires that can withstand large currents and high outputs. As described above, the copper bonding wires for power semiconductors, which are becoming increasingly miniaturized, have the problem of "low followability", and furthermore, due to the limitation of bondability, there are problems with the "cutting property" and "bonding strength" of the wires. Simultaneously solving these three wire problems is the greatest mission of the present invention.

[0021] Patent Document 1 has an object of providing a copper bonding wire with an improved storage life in the atmosphere, and it is described that the storage life can be improved by a copper bonding wire for a semiconductor device having a surface crystal grain boundary density of 0.6 (μm / μm 2 ) or more and 1.6 (μm / μm 2 ) or less, but there is no mention of the problems of the cutting property and followability of the wire.

[0022] Patent Document 2 states that a problem in mass production of the mounting process using copper bonding wires is that the low bonding properties require special bonding conditions, which is a cause of delaying the expansion of practical applications. It also states that this problem can be solved by setting the total proportion of Cu, Cu2O, CuO, and Cu(OH)2 measured by X-ray photoelectron spectroscopy on the wire surface to 10%, and setting the ratio of Cu[II] / Cu[I], which is the total proportion of CuO and Cu(OH)2 corresponding to Cu2 valence (Cu[II]) to the proportion of Cu2O corresponding to Cu1 valence (Cu[I]), to a range of 0.8 to 12. However, it does not mention the issues of wire cutting ability or conformability. Furthermore, Patent Document 2 is an invention of a thin-diameter bonding wire for high-density mounting semiconductor devices, and its application differs from that of a thick-diameter bonding wire for power semiconductors, and the characteristics and properties required of the wire are also different.

[0023] Patent Document 3 aims to provide a Cu alloy bonding wire for semiconductor devices that can meet the performance requirements for high-density LSI applications, wherein the angular difference of the crystal orientation of the wire surface with respect to a direction perpendicular to one plane containing the wire's central axis is 15 degrees or less. <100> Crystal orientation, <110> Crystal orientation, <111> While it is stated that the problem can be solved by having the relative abundance of each crystal orientation be between 3% and 27% in average area fraction, there is no mention of improving cutting performance and bonding strength, or suppressing wire detachment from the tool. Furthermore, the invention described in Patent Document 3, like Patent Document 2, relates to a bonding wire with a thin wire diameter for high-density LSIs, and is fundamentally different from bonding wires for power semiconductor applications. [Prior art documents] [Patent Documents]

[0024] [Patent Document 1] International Publication No. 2021 / 111908 [Patent Document 2] International Publication No. 2021 / 167083 [Patent Document 3] International Publication No. 2020 / 059856 [Overview of the Initiative] [Problems that the invention aims to solve]

[0025] The present invention has been made to solve the above-mentioned problems, and aims to provide a copper bonding wire (hereinafter also referred to as "copper wire") that can improve cutting performance and bonding strength and suppress wire detachment from the tool in wire bonding for connecting, for example, power semiconductors and the busbars of lithium-ion battery mobile devices to the electrodes of lithium-ion batteries. Furthermore, the present invention aims to provide a method for manufacturing a copper wire that can improve cutting performance and bonding strength and suppress wire tool detachment in wire bonding related to the connection of busbars and battery electrodes of power semiconductors and the above-mentioned battery mobile devices, and a semiconductor device using the copper wire. [Means for solving the problem]

[0026] As a result of diligent research, the inventors have found that the copper wires of the following embodiments, the copper wires manufactured by the manufacturing method of the embodiments, and the semiconductor devices can simultaneously solve the problems of cutting performance, bonding strength, and tool detachment of the copper wires during wire bonding.

[0027] The copper bonding wire of the embodiment is made of a copper alloy with a copper purity of 99.99% by mass or higher, and the grain boundary density in a cross-section perpendicular to the wire axis of the copper bonding wire is 0.01 μm. -1 Above 0.6 μm -1 Less than and among the crystal orientations in the wire axis direction in the cross-section, the crystal orientation having an angular difference of 15° or less with respect to the wire axis direction. <111> The orientation ratio of the crystal orientation <101> The value obtained by dividing by the orientation ratio is 10 or more and 650 or less, the dynamic hardness in the cross-section is 45 or more and 90 or less, and the modulus of elasticity in the cross-section is 20 GPa or more and 70 GPa or less.

[0028] Preferably, the copper alloy contains 5 ppm to 30 ppm of silver (Ag) relative to its total weight, and 0 ppm to 3 ppm of phosphorus (P), iron (Fe), silicon (Si), arsenic (As), and antimony (Sb) each, for a total of 15 ppm or less.

[0029] Furthermore, the diameter of the copper wire in the embodiment is preferably 40 μm or more and 700 μm or less.

[0030] Furthermore, the method for manufacturing copper wire in the embodiment is a method for manufacturing copper bonding wire made of a copper alloy with a copper purity of 99.99% by mass or higher, The process involves preparing copper wires made of a copper alloy with a mass of 99.99% or more, The process of drawing the aforementioned copper wire, The process includes a step of applying a final heat treatment to the drawn copper wire, The grain boundary density of the copper bonding wire in a cross-section perpendicular to the wire axis is 0.01 μm. -1 Above 0.6 μm -1 It is less than, Among the crystal orientations in the wire axis direction in the aforementioned cross-section, the crystal orientations whose angular difference with respect to the wire axis direction is 15° or less. <111> The orientation ratio of the crystal orientation <101> The value obtained by dividing by the azimuth ratio is between 10 and 650, The dynamic hardness in the aforementioned cross-section is 45 or more and 90 or less. Furthermore, this is a method for manufacturing a copper bonding wire having an elastic modulus of 20 GPa or more and 70 GPa or less in the cross-section.

[0031] The semiconductor device of the embodiment comprises at least one substrate, at least one semiconductor element on the substrate, an electrode on the surface of the semiconductor element, a semiconductor element substrate on the semiconductor element, and at least one circuit pattern on the substrate. A power semiconductor device comprising electrodes on the surface of the semiconductor element, electrodes on the surface of the semiconductor element and an external electrode, electrodes on the surface of the semiconductor element and one of the circuit patterns, electrodes on the surface of the semiconductor element and a terminal, two adjacent circuit patterns from the circuit patterns, one of the circuit patterns and a terminal, substrates and substrates, and one or more copper wires selected from the group consisting of semiconductor element substrates and circuit patterns on the semiconductor element, The copper wire is a copper bonding wire made of a copper alloy with a copper purity of 99.99% by mass or higher, and the grain boundary density in the cross-section perpendicular to the wire axis of the copper bonding wire is 0.01 μm. -1 Above 0.6 μm -1 Less than and among the crystal orientations in the wire axis direction in the cross-section, the crystal orientation having an angular difference of 15° or less with respect to the wire axis direction. <111> The orientation ratio of the crystal orientation <101> The value obtained by dividing by the orientation ratio is 10 or more and 650 or less, the dynamic hardness in the cross-section is 45 or more and 90 or less, and the modulus of elasticity in the cross-section is 20 GPa or more and 70 GPa or less. In this specification, the symbol "~" represents a numerical range greater than or equal to the value to the left of the symbol and less than or equal to the value to the right of the symbol. [Effects of the Invention]

[0032] According to the copper bonding wire of the present invention, it is possible to improve cutting performance and bonding strength in wire bonding and to suppress wire detachment from the tool. According to the method for manufacturing copper bonding wires of the present invention, for example, in wire bonding for joining busbars and battery electrodes used in power semiconductors and battery mobile devices, it is possible to obtain copper wires that can improve cutting performance and bonding strength, and suppress wire detachment from the tool. According to the semiconductor device using copper wire of the present invention, delamination (lift-off) does not occur at the wire joint, thus providing long-term stability of the joint. Furthermore, due to the excellent conformability of the copper wire, wedge joint defects do not occur, resulting in even greater long-term reliability of the joint (first joint and second joint). [Brief explanation of the drawing]

[0033] [Figure 1] This is a schematic cross-sectional view showing the configuration of a semiconductor device according to the embodiment. [Figure 2] This is a schematic cross-sectional view showing the configuration of a semiconductor device according to another embodiment. [Figure 3] This is a schematic cross-sectional view showing the configuration of a semiconductor device according to another embodiment. [Figure 4] This photograph shows traces of the bonding wire separating from the first joint in a prototype where joint delamination (lift-off) occurred. [Figure 5] This photograph shows traces of the bonding wire separating from the second joint in a prototype where joint delamination (lift-off) occurred. [Figure 6] This photograph shows the state of the bonder when there is leftover wire after the second joining. [Figure 7] This is a cross-sectional photograph of a copper wire in an embodiment where the grain boundary density is 0.18 μm-1. [Figure 8] This is a photograph of a cross-section of a copper wire with a grain boundary density of 0.68 μm⁻¹. [Figure 9] This is a photograph of a wire with no tool detachment. [Figure 10] This is a photograph of a wire that has come loose from the tool. [Figure 11] This is a schematic diagram illustrating the structure of the bond head and its state immediately after wire joining in a front-cut method. [Figure 12] This is a schematic diagram illustrating the structure of the bond head and its state immediately after wire joining in a rear-cut method. [Figure 13]It is a diagram schematically showing a part of a lithium-ion battery module. [Figure 14] It is a diagram schematically showing a part of a lithium-ion battery module of another aspect. [Figure 15] It is a diagram schematically showing the whole lithium-ion battery module.

Mode for Carrying Out the Invention

[0034] Hereinafter, the copper bonding wire of the embodiment of the present invention will be described. The copper wire of this embodiment is suitably used for power semiconductors and for connecting bus bars and battery electrodes of the battery mobile. The copper wire of this embodiment is made of a copper alloy with a copper purity of 99.99 mass% or more, and the grain boundary density in the cross-section perpendicular to the wire axis of the copper wire is 0.01 μm -1 or more and less than 0.6 μm -1 and, among the crystal orientations in the wire axis direction in the cross-section, the value obtained by dividing the orientation ratio of the crystal orientation <111> with an angular difference of 15° or less with respect to the wire axis direction by the orientation ratio of the crystal orientation <101> is 10 or more and 650 or less, the dynamic hardness in the cross-section is 45 or more and 90 or less, and the elastic modulus in the cross-section is 20 GPa or more and 70 GPa or less. Hereinafter, the process of trial and error until the present invention is reached, and the configuration and manufacturing method of the copper wire of this embodiment will be described in detail.

[0035] As wire cut error, there is peeling (lift-off) of the wire from the joint between the wire and the substrate or the like before the wire itself breaks. Lift-off is caused by the joint strength being weaker than the breaking strength of the wire. As factors that greatly affect lift-off, in addition to cutting performance, there is also the weakness of the joint strength itself at the joint between the wire and the substrate or the like. A method for improving the joint strength will be described later.

[0036] First, the inventors carefully observed numerous prototypes produced under various manufacturing conditions to investigate the relationship between the cutting properties of copper wire and the microstructure of the cross-section perpendicular to the wire axis. As a result, they found a correlation between the density of grain boundaries and cutting properties. Hereafter, the cross-section perpendicular to the wire axis of the copper wire will also be referred to as the "transverse plane."

[0037] Specifically, to evaluate the continuous bonding performance of the numerous prototypes mentioned above, continuous bonding was performed using a wire bonding apparatus. The prototypes were divided into those that bonded without stopping until completion and those that stopped midway. A detailed investigation of the causes of the prototypes that stopped midway revealed two main causes: stopping due to joint delamination (lift-off) and stopping because the wire was pulled up to the set height and attempted to be torn off but failed to cut (stopping due to exceeding the height limit). Figure 4 is a photograph showing the traces of the bonding wire detaching from the first joint in a prototype where joint delamination (lift-off) occurred. Figure 5 is a photograph showing the traces of the bonding wire detaching from the second joint in a prototype where joint delamination (lift-off) occurred. Incidentally, in Figure 5, the scratches of a predetermined width running vertically across the paper are the marks of the cutter blade. When cutting the wire using the front-cut method (full-cut method), there is a possibility of damaging the bonding mating material if it is delicate. Figure 6 is a photograph showing the bonder's behavior when there is remaining wire after the second joint. From the situation in Figure 6, it can be inferred that the bonder stopped due to exceeding the height limit.

[0038] The inventor carefully observed the crystalline structure of the cross-section of each prototype wire used in continuous bonding tests and noticed that the density of the grain boundaries affected the cutting properties of the wire, which in turn greatly affected the continuous bonding properties. Specifically, grain boundary density = grain boundary length (μm) / wire cross-sectional area (μm) 2 )) is 0.01 μm -1 Above 0.6 μm -1 If the thickness is less than 0.01 μm, the wire bonding apparatus can perform continuous bonding without stopping until the end.-1 If it is less than 0.6μm, the wire will stretch too much before it is cut, causing it to stop due to exceeding the height limit. -1 We found that, under these conditions, the wire tends to stop due to delamination (lift-off) from the joint before it is cut.

[0039] Generally, the smaller the crystal grain size, the longer the total length of the grain boundary tends to be, and the larger the crystal grain size, the shorter the total length of the grain boundary tends to be. However, individual crystal grain shapes are diverse, and even with nearly identical crystal grain sizes, the total length of the grain boundary differs depending on the crystal grain shape. For example, in the case of polygonal crystal grains, even with nearly identical crystal grain sizes, the grain boundary length differs greatly depending on the number of sides of the polygon. Therefore, the inventors devised that using the grain boundary density as an indicator is more appropriate than using the crystal grain size when investigating the relationship with cutting properties. Figure 7 shows a crystal grain boundary density of 0.18 μm. -1 Figure 8 is a cross-sectional photograph of the copper wire of the embodiment, where the grain boundary density is 0.68 μm. -1 This is a photograph of a cross-section of a copper wire.

[0040] Furthermore, the length of the grain boundaries also differs depending on the wire diameter. Basically, the thicker the wire, the longer the grain boundaries, and the thinner the wire, the shorter the grain boundaries. Therefore, in this invention, the total length of the grain boundaries (μm) is defined as the area of ​​the cross-section perpendicular to the longitudinal direction of the wire (μm). 2 We decided to use the grain boundary density value obtained by dividing by ). The method for measuring the grain boundary density will be described later.

[0041] The grain boundary density in the cross-section of the copper bonding wire in this embodiment is 0.01 μm -1 Above 0.6 μm -1 It is less than . This can improve the cutting performance of the copper wire. The grain boundary density is more preferably 0.03 μm. -1 Above 0.5 μm -1 The following, and more preferably 0.05 μm -1 More than 0.3μm -1 The following applies:

[0042] Although the relationship between cutting ability and grain boundary density, and the details of the mechanism, are unclear, it is presumed that when a wire is torn apart after a crack has been induced, a lower grain boundary density makes it easier for crack propagation to concentrate at certain grain boundaries, and this contributes to the ease with which the wire can be cut.

[0043] Next, let's discuss bonding strength. While aluminum has traditionally been used for bonding wires in power semiconductors, copper wires are harder and less prone to deformation than aluminum wires. As a result, the bonding area at the joint tends to be smaller in wedge bonding, which leads to a decrease in bonding strength. If a large load is applied during bonding to forcibly deform the wire in order to increase the bonding area, there is a high possibility of damaging the bonded object, such as a delicate semiconductor chip.

[0044] The more easily a copper wire deforms, the easier it is to crush, allowing for a larger bonding area even with a small load. This increases the bonding strength and reduces the likelihood of chip damage. After investigating various indicators of crushability, the inventor concluded that dynamic stiffness, which takes into account the elastic properties of the wire, is the optimal indicator.

[0045] Vickers hardness (Hv) and Knoop hardness (HK), widely used for measuring the hardness of metallic materials, are methods that involve contacting an indenter with the sample surface, applying a test force to create a depression, and then calculating the surface area from the diagonal length of the depression after removing the load to determine the hardness. However, this method cannot evaluate properties that reflect the elasticity of the material, and therefore, when considering the bonding process in which a wire is pressed down with a wedge tool and joined by applying ultrasound while elastic deformation is applied, we considered it unsuitable as an indicator of wire properties.

[0046] Vickers hardness is the most common measure of hardness. It is a method of evaluating hardness by pressing a diamond indenter into a material and measuring the shape (area) of the indenter left on the material. In Vickers hardness, the indenter is pulled back after being pressed into the material, and the material returns slightly due to its elastic properties. As a result, the area of ​​the indentation left by the indenter becomes slightly smaller, and the measured hardness value is slightly higher than the actual hardness.

[0047] Dynamic hardness is a value obtained from the test force and indentation depth during the process of pressing the indenter in, and can reflect not only the plasticity of the sample but also the elastic rebound deformation. The inventors believed that dynamic hardness is very suitable as a value representing the properties of the copper wire in the embodiment because the series of actions of dynamic hardness measurement are similar to the action of a wedge tool gripping a wire and then crushing and deforming the wire. Furthermore, since dynamic hardness is measured with the indenter pressed into the material, there is no influence from the elastic rebound of the material. Therefore, the inventors believed that dynamic hardness is an optimal evaluation method for representing the properties of materials with a high elastic modulus, especially copper wire.

[0048] Furthermore, when measuring dynamic hardness, the elastic modulus of the wire can also be measured by applying and removing the test force. The inventor considered the elastic modulus to be an important parameter indicating the wire's ease of deformation and conformability, and therefore included it as an evaluation item.

[0049] The inventors measured the dynamic hardness and elastic modulus of many prototypes manufactured using different methods and found that when the dynamic hardness of the copper wire in its cross-section is between 45 and 90, and the elastic modulus of the copper wire in its cross-section is between 20 GPa and 70 GPa, the wire can be crushed with a small load without causing any problems. This allows the bonding area of ​​the joint to be increased, thus resolving the issue of insufficient bonding strength.

[0050] The dynamic hardness (DH) of the copper wire in this embodiment is preferably between 50 and 80, and even better between 55 and 70. If the dynamic hardness is less than 45, when the wire is crushed, the bonding area expands too much, increasing the likelihood of contact with adjacent wires and causing a short circuit. If it exceeds 90, the wire's strength is too great, making it difficult to crush, which not only prevents increasing the bonding area but also increases the likelihood of damaging the semiconductor chip when bonding.

[0051] The modulus of elasticity in the cross-section of the copper wire in this embodiment is preferably 30 GPa to 60 GPa, and even better, 35 GPa to 50 GPa. If the modulus of elasticity is less than 20 GPa, there is a high possibility of a leaning failure occurring, where the wire joined from the first joint to the second joint collapses. If it exceeds 70 GPa, there is a high possibility of the wire detaching from the wedge tool, as described later. It is desirable to measure the dynamic hardness and modulus of elasticity in the region on the outer circumference of the cross-section perpendicular to the longitudinal direction of the wire, rather than the midpoint of the radius.

[0052] Furthermore, dynamic hardness and elastic modulus can be measured, for example, using a micro-compression testing machine MCT-W500 manufactured by Shimadzu Corporation, with a triangular pyramidal indenter having a ridge angle of 115°, under the following conditions. Test mode: Load-unload test Test force: 980.665 mN Load speed: 20.7411mN / sec Hold time: 10 seconds

[0053] The bonding strength was measured as follows: A HESSE BJ935 bonder was used to bond wires to a copper plate measuring 50mm (length) x 50mm (width) x 1mm (thickness), and then a pull test was performed. The bonding conditions were set to a load of 2500gf and a power of 50V. The lift-off rate of the joint was compared for each prototype when a pull test was performed on the center of the loop using 20 prototype wires (n=20). In the n=20 pull test, a lift-off of 3 or more times was considered a failure, and a lift-off of less than 3 times was considered a pass.

[0054] Next, we will discuss further challenges to bonding wires associated with the miniaturization of power semiconductors. Normally, when there is sufficient space for wire bonding, the bonding operation can be performed in a straight line, and the wire loop curve is not complicated. However, when there are limitations on the area for bonding, bonding the required number of wires within a limited area results in a complex loop curve, as well as bonding operations involving sharp bends and twists. For example, in the first bonding (connection to electrodes on the semiconductor chip) and second bonding (connection to external electrodes on the lead frame or substrate) of copper wires, a wedge tool (sometimes simply called a "tool") with alligator clips (grooves) at its tip is used to hold the wire. The wire is then fitted into these alligator clips and pressed against the bonding site to perform the bonding. In bonding operations with many bends and twists, the wire held by the wedge tool (clamp) may not be able to follow the movement and may slip out of the alligator clips of the tool. Copper wires, which have a higher modulus of elasticity (springiness) than conventionally used aluminum wires, are overwhelmingly more prone to tool slippage.

[0055] If a portion of the wire is detached from the wedge tool during the second bonding process, sufficient bonding strength may not be ensured, and depending on the operating environment, the bond may peel off, potentially leading to failure. Furthermore, if the second bonding is performed on a chip with the wire detached from the tool, the tip of the wedge tool may come into direct contact with the metal film on the element, potentially damaging the semiconductor element.

[0056] By controlling the range of the elastic modulus in the cross-section of the copper wire to between 20 GPa and 70 GPa, some degree of tool slippage was eliminated. However, in order to further eliminate tool slippage, the inventors diligently studied and found that, in addition to the elastic modulus, among the crystal orientations in the direction of the wire axis in a cross-section perpendicular to the wire axis, the angular difference with respect to the wire axis is 15° or less. <111> The orientation ratio of the crystal orientation <101> We discovered that by controlling the value obtained by dividing by the azimuth ratio to be between 10 and 650, most tool detachment problems can be eliminated. Figure 9 is a photograph of a wire without tool detachment. In the photograph in Figure 9, wire 91 is fitted into the alligator clip 90. Figure 10 is a photograph of a wire with tool detachment. In the photograph in Figure 10, wire 92 has detached from the alligator clip 90.

[0057] Specifically, in a cross-section perpendicular to the wire axis, the crystal orientations in the direction of the wire axis are those whose angular difference with respect to the wire axis is 15° or less. <111> The orientation ratio of the crystal orientation at the same location and under the same conditions as above <101> The inventors discovered that by dividing the value by the orientation ratio (hereinafter also referred to as the "crystal orientation ratio") between 10 and 650, most tool slippages are eliminated. The crystal orientation ratio in the cross-section perpendicular to the wire axis is better between 20 and 500, and even better between 30 and 450. If the crystal orientation ratio is less than 10, tool slippage is less likely to occur, and if it exceeds 650, tool slippage will not occur, but the possibility of other problems such as inability to cut properly increases. The crystal orientation ratio can be calculated using the following formula. Crystal orientation ratio ratio = (crystal orientation <111> (orientation ratio) / (crystal orientation) <101> (Directional ratio)

[0058] The mechanism by which a crystal orientation ratio ratio in a cross-section perpendicular to the wire axis is between 10 and 650 is effective in preventing tool slippage is not clear, but the crystal orientation <111> This refers to the orientation of the slip plane in wire EDM, and it is presumed that the more of these crystal orientations there are, the more flexibly the tool can respond to sudden movements. Also, there are crystal orientations other than the crystal orientation of the slip plane. <101> The inventors succeeded in discovering that by determining the ratio of this, the wire's ability to follow the tool's movement can be improved. Incidentally, the crystal orientation <111> and crystal orientation <100> We also examined the crystal orientation ratio ratio, but no particularly distinctive trends were observed.

[0059] Incidentally, copper oxidizes easily, and if a copper oxide film forms on the surface of a copper wire, the bonding strength weakens.

[0060] Furthermore, as mentioned above, the copper wire of the present invention has a diameter of 40 μm to 700 μm, which is thicker than the wires commonly used in semiconductor devices (diameter 15 μm to 30 μm), and also has a larger bonding area (volume). Therefore, it can carry a larger voltage and a larger current compared to thinner wires. However, if the electrical resistance of the copper wire itself is high, the current will be smaller even at the same voltage, and the advantage of its thickness will be lost. The inventors considered the electrical conductivity of ultra-high-purity copper with few impurities (99.99% by mass) to be 100%, and aimed to ensure that the copper wire of this embodiment (copper wire with a purity of 99.99% by mass or higher), which contains more impurities, also has an electrical conductivity of at least 99% when the electrical conductivity of ultra-high-purity copper is set to 100%.

[0061] As mentioned earlier, the widespread adoption of electric vehicles (EVs) is expected to be one way to combat global warming (reduce carbon dioxide emissions). However, for EVs to become widespread, it is desirable to solve the problem of EVs having a shorter driving range on a full charge compared to gasoline cars. To extend the driving range of EVs, miniaturization (lightening) of EVs is a crucial issue. On the other hand, the greater the energy loss and heat dissipation caused by the electrical resistance of copper wires during the switching (on / off) of power semiconductors, the more space is required for cooling, resulting in larger EVs. One of the obstacles to miniaturizing automobiles is this electrical resistance of copper wires; in other words, lowering the electrical resistance of the wires is a key to promoting the widespread adoption of EVs.

[0062] The inventors investigated how to simultaneously solve the problems of oxide formation on the wire surface and increased electrical resistance by adjusting the composition of a copper alloy wire. They sought alloying elements that would suppress the progression of oxide formation and the increase in electrical resistance in 99.99% pure copper. After diligently conducting research experiments by adding numerous elements to a copper base material, they discovered that adding silver (Ag) and, in particular, controlling the content of phosphorus (P), iron (Fe), silicon (Si), arsenic (As), and antimony (Sb) could suppress copper oxide formation and the increase in electrical resistance.

[0063] As a result of numerous experiments, the inventors found that the amount of silver (Ag) relative to the total amount of copper alloy is preferably between 5 ppm and 30 ppm by mass. Furthermore, while phosphorus (P), iron (Fe), silicon (Si), arsenic (As), and antimony (Sb) do not necessarily have to be included, it was found that it is preferable to adjust the content of each element to 3 ppm by mass or less, and that the total amount of phosphorus (P), iron (Fe), silicon (Si), arsenic (As), and antimony (Sb) relative to the entire copper wire is preferably 15 ppm by mass or less.

[0064] The inventors discovered that when the silver (Ag) content is between 5 ppm and 30 ppm relative to the total amount of copper alloy, and at the same time, when the phosphorus (P), iron (Fe), silicon (Si), arsenic (As), and antimony (Sb) content is between 0 ppm and 3 ppm for each element, and the total is 15 ppm or less, it is easy to obtain an effect that delays the formation of copper oxide and prevents the overall electrical resistance of the copper wire from increasing too much.

[0065] In particular, in the manufacturing process of the copper wire of this embodiment, when the copper wire is exposed to high temperatures during heat treatment, the rate of copper oxide formation increases, depending on the conditions of the heat treatment atmosphere. In this case, the copper wire controlled to the above composition has the effect of slowing down the rate of copper oxide formation.

[0066] The progression of oxidation on the surface of a copper wire can be evaluated by measuring the copper(II) oxide (CuO) film formed on the copper wire surface after a standing test. The standing test involves leaving the copper wire sample unsealed in a container at room temperature (35°C) and humidity (75% RH) for up to 10 months. During this time, the CuO oxide film formed on the copper ribbon surface is measured at 1 hour, 1 month, 2 months, 3 months, and so on, every month until 10 months. In the process of copper oxidation, copper(I) oxide (Cu2O) is formed first, and as oxidation progresses, the proportion of CuO increases. Therefore, by calculating the proportion of the CuO oxide film, the ease of oxidation formation (oxidation resistance) of the copper wire surface can be evaluated. A larger proportion of the CuO oxide film indicates that oxidation has progressed further. The proportion of the CuO oxide film can be calculated using the following formula.

[0067] Percentage of CuO oxide film = ((CuO oxide film) / (CuO oxide film + Cu2O oxide film)) × 100 (%)

[0068] In this embodiment, the percentage of the CuO oxide film after 6 months is used as the evaluation criterion. If the percentage of copper(II) oxide (CuO) film is 30% or less, it is evaluated as a pass ("A"), and if it is 20% or less, it is evaluated as a better ("S"). The analytical method will be described later.

[0069] Furthermore, the International Annealed Copper Standard (IACS) was used for electrical conductivity. Compared to the IACS of 99.999 mass% pure copper (100%), samples with an IACS decrease of less than 0.5% were designated as "S" to indicate excellent performance. Samples with an IACS decrease of more than 0.5% but less than 1% were designated as "A" to indicate good performance. Details of the IACS measurement method will be described later.

[0070] The diameter of the copper wire in this embodiment is typically 40 μm to 700 μm, preferably 70 μm to 600 μm, and more preferably 100 μm to 500 μm.

[0071] (Method of manufacturing copper wire) Next, an example of a method for manufacturing the copper wire of the embodiment will be described. Note that the method for manufacturing the copper wire is not limited to the method shown below. Furthermore, it is desirable to adjust the conditions as appropriate, taking into account the weight of the copper wire to be manufactured and the processing capacity of the heat treatment furnace.

[0072] High-purity copper with a purity of 99.99% by mass or higher is dissolved together with silver, depending on the composition of the copper wire, to produce molten copper. Heating furnaces such as arc heating furnaces, high-frequency heating furnaces, resistance heating furnaces, and continuous casting furnaces are used for melting. While melting in air is acceptable, it is desirable to melt the molten copper in the heating furnace under vacuum or in an inert gas atmosphere such as argon or nitrogen to prevent the incorporation of air from the atmosphere. The molten material is solidified from the heating furnace by continuous casting to produce wire strands of the desired diameter. Alternatively, molten copper can be poured into a mold to create an ingot, which can then be set in an extruder and extruded to the desired wire diameter. Furthermore, rough-drawn wire can be produced using the DIP forming method or SCR method and used as wire strands.

[0073] The wires obtained in the above process are drawn into intermediate wires with a diameter of 900 μm. Next, the intermediate wires are subjected to an intermediate heat treatment by heating at 400°C to 600°C for 60 to 420 minutes. The intermediate heat treatment can be carried out in a "combustion furnace" using petroleum or gas as a heat source, or in an "electric furnace" using electrical energy as a heat source, and can be done in a "batch type" or "continuous type". Alternatively, the wire can be heated by direct current, or in an "electrical type" method. From the viewpoint of preventing oxidation of the wire, an inert gas atmosphere is desirable for the heat treatment atmosphere.

[0074] Next, the wire that has undergone intermediate heat treatment is drawn to the target final wire diameter. In wire drawing, the wire is passed through multiple carbide dies or diamond dies in sequence, gradually reducing the wire diameter. The wire reduction rate (also called the surface area reduction rate or processing rate) before and after each die drawing process should ideally be between 5% and 30%.

[0075] As a finishing touch, the wire processed to its final diameter undergoes heat treatment (final heat treatment). The conditions for the final heat treatment are the most important part of manufacturing the copper wire of this embodiment. After conducting various heat treatment experiments, the inventors determined that the following method is the most preferable for manufacturing the copper wire of this embodiment. An electric heating method was adopted for the final heat treatment. The conditions for electric heating were a voltage of 16V to 28V, a distance between electrode terminals of 800mm to 1300mm, a wire travel speed of 50m / min to 200m / min, and the second electrode was immersed in cooling water (i.e., the wire moved into the cooling water while being heated). Pure water was used as the cooling liquid, with a pure water temperature of 20°C to 80°C and a dissolved oxygen concentration of 8mg / liter or less. In addition, the space between the electrode terminals through which the wire passes was covered with a cylindrical structure to prevent wire oxidation, and nitrogen gas was flowed at a flow rate of 30 liters / min to 60 liters / min. Although the electric heating method has been described here, other heat treatment methods may be used under equivalent heating conditions.

[0076] (Semiconductor device) Next, the configuration of the semiconductor device 100 using copper wires according to the embodiment will be described with reference to Figure 1.

[0077] As shown in Figure 1, the semiconductor device 100 comprises a semiconductor element 1, a metal film 2, a wire 3, a circuit pattern 41, a metal pattern 42, an insulating member 43, a heat dissipation member 5, a bonding material 6, a case 7, terminals 8, and a sealing material 9.

[0078] In this embodiment, the semiconductor element 1 is, for example, a power semiconductor used in a power supply semiconductor. Examples of semiconductor elements 1 include metal oxide semiconductor field-effect transistors (MOSFETs) and insulated gate bipolar transistors (IGBTs).

[0079] The semiconductor device 1 is formed by stacking electrodes 11, a substrate portion 13, and a back electrode 12 in this order. The electrodes 11 are, for example, aluminum (Al)-silicon (Si) electrodes, and the substrate portion 13 are, for example, a silicon (Si) substrate, a silicon carbide (SiC) substrate, a gallium nitride (GaN) substrate, etc.

[0080] The metal film 2 is provided on the surface of the electrode 11 opposite to the substrate portion 13, so as to cover the surface of the electrode 11. The metal film 2 is a nickel (Ni) film, a copper (Cu) film, a titanium (Ti) film, a tungsten (W) film, etc., and is a film formed by electroplating, electroless plating, vapor deposition, sputtering, etc. Nickel (Ni) films include nickel (Ni) electroless plated films, specifically electroless nickel (Ni)-phosphorus (P) plated films, electroless nickel (Ni)-boron (B) plated films, etc. Other preferred embodiments of the metal film 2 will be described later.

[0081] Wire 3 consists of the copper wire of the embodiment described above, and its structure and characteristics are also as described above. Wire 3 is bonded to the surface of the metal film 2.

[0082] Next, the other components of the semiconductor device 100 will be described. Inside the semiconductor device 100, a semiconductor circuit is formed by semiconductor elements 1, wires 3, terminals 8, circuit patterns 41, and metal patterns 42. Inside the semiconductor device 100, wires 3 are bent, and this bent portion is used to join the semiconductor elements 1, terminals 8, circuit patterns 41, etc.

[0083] In the semiconductor device 100, a bonding material 6, a metal pattern 42, an insulating material 43, a circuit pattern 41, another bonding material 6, and a semiconductor element 1 are stacked in that order on the surface of the heat dissipation member 5. The bonding material 6 consists of solder, silver (Ag), etc., and is used to bond the heat dissipation member 5 to the metal pattern 42, and the circuit pattern 41 to the back electrode 12 of the semiconductor element 1, respectively. The insulating material 43 is an insulating substrate or the like.

[0084] Case 7 consists of a ring-shaped enclosure with an internal space and is provided so as to surround the outer circumference of the heat dissipation member 5. The semiconductor element 1, metal film 2, wire 3, circuit pattern 41, metal pattern 42, insulating member 43, bonding material 6, and sealing material 9 described above are housed in the internal space of Case 7.

[0085] Terminal 8 functions as a connection terminal for external equipment. Terminal 8 is provided on the top surface of case 7, with one end protruding into the internal space of case 7 and the other end protruding into the external area of ​​case 7. The sealing material 9 is filled into the internal space of case 7, enclosing the semiconductor element 1, metal film 2, wire 3, circuit pattern 41, metal pattern 42, insulating member 43, and bonding material 6. The sealing material 9 is a gel-like sealing resin or a cured molding resin.

[0086] The semiconductor device 100 shown in Figure 1 may have electrodes 11 on a plurality of semiconductor elements 1, a plurality of circuit patterns 41, and a plurality of terminals 8. Preferably, the semiconductor device 100 has an electrode-circuit pattern junction structure including one electrode 11 from the plurality of semiconductor elements 1, one circuit pattern 41 from the plurality of circuit patterns 41, and a wire 3 connecting the electrode 11 and the circuit pattern 41. Preferably, the semiconductor device 100 also has an electrode-terminal connection structure including one terminal 8 from the plurality of terminals 8, one electrode 11 from the plurality of electrodes 11, and a wire 3 connecting the terminal 8 and the electrode 11. Furthermore, preferably, the semiconductor device 100 has a circuit pattern-terminal connection structure including one terminal 8 from the plurality of terminals 8, one circuit pattern 41 from the plurality of circuit patterns 41, and a wire 3 connecting the terminal 8 and the circuit pattern 41. Preferably, the semiconductor device 100 includes one or more electrode-circuit pattern junction structures, electrode-terminal connection structures, and circuit pattern-terminal connection structures, and more than one of each. In other words, the copper wire of the embodiment can be used for connecting the electrode 11 to the circuit pattern 41, for connecting the terminal 8 to the electrode 11, or for connecting the terminal 8 to the circuit pattern 41.

[0087] The semiconductor device 100 may also have a substrate (a substrate on the semiconductor element) on top of the semiconductor element 1. In this case, it is preferable that the semiconductor device 100 has a junction structure that includes the substrate on the semiconductor element, a circuit pattern 41, and wires 3 connecting the substrate on the semiconductor element and the circuit pattern 41. The same applies to the semiconductor devices 101 and 103 described later.

[0088] Figure 2 shows another embodiment of a semiconductor device, a semiconductor device 101 having a lead frame. In Figure 2, components that perform the same function as the semiconductor device 100 shown in Figure 1 are denoted by the same reference numerals, and detailed explanations are omitted. The semiconductor device 101 shown in Figure 2 has a lead frame LF in addition to a semiconductor element 1, a metal film 2, wires 3, an insulating member 43, a bonding material 6, and a sealing material 9. The semiconductor device 101 shown in Figure 2 does not have a case 7 because it has a lead frame LF, but it may have a case 7. The lead frame LF is bonded to the surface of the insulating member 43 and has the same function as the circuit pattern 41 of the semiconductor device 100 shown in Figure 1. In Figure 2, the lead frame LF and the insulating member 43 are bonded together, but a metal plate (not shown) may be placed between the lead frame LF and the insulating member 43.

[0089] The encapsulating material 9 is provided to enclose the semiconductor element 1, metal film 2, wire 3, insulating member 43, bonding material 6, and lead frame LF. However, the ends of the lead frame LF protrude outside the encapsulating material 9, and the lead frame LF constitutes the electrical circuit of the semiconductor element 1 and wire 3, with the protruding ends functioning as terminals 8 for connecting to external equipment of the semiconductor device 101.

[0090] Figure 3 schematically shows another embodiment of a semiconductor device, a semiconductor device 103 having a circuit pattern 41 and a wire 3 connecting the circuit patterns 41. The semiconductor device 103 differs from the semiconductor device 100 shown in Figure 1 in that it has another circuit pattern 41 next to the circuit pattern 41 of the semiconductor device 100, and a wire 3 connects adjacent circuit patterns 41, but the other configurations are the same. The semiconductor device 103 shown in Figure 3 may have a plurality of circuit patterns 41, and includes a junction structure comprising a wire 3 connecting two adjacent circuit patterns 41. The semiconductor device 103 includes one or more of these junction structures, preferably two or more. That is, the copper wire in the embodiment can be used for connecting circuit patterns 41 to circuit patterns 41.

[0091] Next, the manufacturing methods for semiconductor devices 100, 101, and 103 shown in Figures 1, 2, and 3 will be described. First, each component constituting semiconductor devices 100 and 101 is prepared, stacked according to the above configuration, and joined together. Then, the ends of the wires 3 are joined to the surface of the metal film 2 by ultrasonic bonding or the like. After that, the other end of the wire 3 is wedge-bonded to an external electrode (terminal 8 in Figure 1 or LF in Figure 2). The copper wire of the above-described embodiment is used as the wire 3. Then, the sealing resin is injected into the semiconductor device 100 and cured to form the sealing material 9. In the case of semiconductor device 101, a lead frame on which the semiconductor elements 1 etc. are mounted is placed in a mold, the sealing resin is injected and cured to form the sealing material 9.

[0092] In the semiconductor device equipped with the copper wire of the above-described embodiment, delamination (lift-off) does not occur at the wire joint, so the joint 31 can be maintained stably for a long period of time. Furthermore, due to the good conformability of the copper wire, wedge joint defects do not occur, making it possible to manufacture a semiconductor device with long-term reliability in the joints (first joint and second joint). [Examples]

[0093] Examples are described below. The present invention is not limited to the following examples.

[0094] The copper wires used in the examples were manufactured by the following method and under the following conditions: Copper with a purity of 99.99% by mass or higher was melted together with silver as an additive element. A continuous casting furnace (heating furnace) was used for melting. To prevent contamination from oxygen and other elements from the atmosphere, the molten copper in the heating furnace was evacuated and then melted under a nitrogen atmosphere. The molten material was solidified by continuous casting from the heating furnace to a predetermined diameter to produce individual wires.

[0095] The wire obtained above was drawn to a diameter of 0.9 mm and subjected to intermediate heat treatment at 500°C for 120 minutes. After the intermediate heat treatment, it was drawn again to a final wire diameter of 0.5 mm (500 μm). The wire drawing process was carried out using multiple carbide dies or diamond dies to gradually reduce the wire diameter. The reduction in surface area before and after each wire drawing process was in the range of 5 to 30%.

[0096] The wire, processed to its final diameter, was placed in an electrically heated furnace for final heat treatment. The conditions for electric heating were: voltage between 16V and 28V, electrode terminal distance between 800mm and 1300mm, wire travel speed between 50m / min and 200m / min, and the second electrode was immersed in cooling water. This ensured that the wire was immersed in cooling water while heated. Pure water was used as the coolant, with a pure water temperature between 20°C and 80°C and a dissolved oxygen concentration of 8mg / L or less during electric heating. The area around the electrodes through which the wire passed was maintained in a nitrogen atmosphere. Samples for Examples 1 to 30 were obtained by varying the conditions within the above range. After the final heat treatment, the wire was rewound onto several spools in units of approximately 200m. The copper purity and some of the contained elements of the copper wire for each example are shown in Table 2.

[0097] Table 2 shows the copper purity and some of the elemental content of the comparative example copper wires. Regarding the manufacturing method of the comparative example copper wires, the manufacturing process from melting to the final wire diameter was carried out under almost the same conditions as the manufacturing method of the example copper wire, but the final heat treatment conditions were changed by using all or part of the above-described range of electrical heating conditions outside the range of the manufacturing conditions of the example. This yielded comparative examples 1 to 10.

[0098] Next, the methods for measuring the characteristics of the copper wires in the examples and comparative examples will be described. (Measurement of grain boundary density and crystal orientation ratio) The crystal orientation of the cross-section perpendicular to the wire axis (longitudinal direction of wire drawing) was measured as follows. Multiple wire samples were prepared by cutting the fabricated copper wire to a length of several centimeters. Care was taken to prevent deformation of the wire samples, and they were attached straight and flat onto an Ag-plated metal frame (metal plate). Then, the wire samples, along with the metal plate, were placed in a cylindrical mold so that the metal plate became the bottom surface of the cylinder, and resin was poured into the mold. After that, a hardening agent was added and the resin was cured. Subsequently, the cylindrical resin containing the cured wire samples was roughly polished with a polisher so that the cross-section perpendicular to the longitudinal direction of the wire was exposed. After that, the cut surface was finished by final polishing, and then residual strain on the polished surface was removed by ion milling to obtain a smooth surface. The ion milling apparatus was finely adjusted so that the wire cut surface was perpendicular to the longitudinal direction of the wire.

[0099] The grain boundary density and crystal orientation ratio of the copper wires in the examples and comparative examples were measured and calculated using the Electron Backscattered Diffraction Pattern (EBSD) method. Specifically, the wire sample was attached to the sample stage of a Field Emission Scanning Electron Microscope (FE-SEM) equipped with an EBSD measuring device (EDAX / TSL OIM Data Collection) so that its cross-section (i.e., the polished surface of the sample) was parallel to the sample stage. The settings were as follows: acceleration voltage 15kV, measurement point spacing 1.5μm, magnification 170x, and boundaries where the orientation difference between adjacent measurement points was greater than 15° were considered grain boundaries. If five or more pixels were connected with an orientation difference of 15° or less, it was recognized as a single crystal grain.

[0100] The grain boundary density is calculated by comparing the total length of the grain boundaries (μm) with the cross-sectional area (μm) of the analyzed area. 2 Divide by ) to obtain the grain boundary density (μm ―1 The grain boundary density was calculated by selecting the vicinity of the tip and trailing end of a wire approximately 200m long, as well as the intermediate section between them. The grain boundary density was measured in cross-sections perpendicular to the longitudinal direction at these three locations, and the average values ​​are shown in Table 1. Similarly, the crystal orientation ratio is also the crystal orientation in the wire cross-section described above, where the angular difference with respect to the wire axis is 15° or less. <111> The orientation ratio of the crystal orientation at the same location and under the same conditions as above <101> The value was calculated by dividing by the orientation ratio. Three measurements were taken at the same locations as the grain boundary density, and the average value is shown in Table 1.

[0101] (Method for measuring dynamic hardness and elastic modulus) Dynamic hardness and elastic modulus were measured using a Shimadzu Corporation micro-compression testing machine MCT-W500 with a triangular pyramidal indenter with a ridge angle of 115°, under the following conditions. Test mode: Load-unload test Test force: 980.665 mN Load speed: 20.7411mN / sec Hold time: 10 seconds For the measurement samples, each 200m wire sampled after the final heat treatment was cut at three points: the tip, the trailing end, and approximately midway between the tip and trailing end. The samples were embedded in resin so that the cross-section perpendicular to the wire axis was exposed approximately perpendicular to the wire axis, and after polishing, measurements were taken at four points near the midpoint of the radius of the wire cross-section. The average of the measurements taken at a total of 12 points is shown in Table 1. The four lines connecting each of the four points to the wire center are at 90-degree angles to each other, and the four points are arranged at approximately equal intervals. Note that the dynamic hardness and elastic modulus are measured almost simultaneously in a series of steps using the above measuring device. However, dynamic hardness is determined by the load on the indenter (depth in the depth direction), and elastic modulus is determined by the return of the indented sample after the load on the indenter is removed. Therefore, each is an independent value specific to the material.

[0102] Next, we will explain the evaluation methods for each sample in the examples and comparative examples. (Method for evaluating cutting performance) The HESSE BJ935 bonder was used to continuously bond wires to a copper plate measuring 50mm x 50mm x 1mm thick, and the device was evaluated based on the number of errors (stops) it generated. The bonding conditions were a load of 2500gf and a power of 50V. For each sample, the combination of the first and second bonding was set to n=1 (1 pair), and continuous bonding was performed for 1000 pairs (n=1000). If the number of errors (stops) was 3 or more, it was evaluated as a failure ("X"). If it was 2, there was room for improvement, but it was evaluated as unlikely to be a practical problem ("B"). If it was 1, it was evaluated as good ("A"), and if it was 0, it was evaluated as excellent ("S"). The results are shown in Table 1.

[0103] (Method for evaluating joint strength) For joint strength testing, a HESSE BJ935 bonder was used, and a pull test was performed after bonding wires to a copper plate measuring 50mm (length) x 50mm (width) x 1mm (thickness). The bonding conditions were set to a load of 2500gf and a power of 50V. One pair of first and second joints was defined as n=1, and the number of joint lift-offs during pull tests at the center of the loop was compared for 20 pairs (n=20). In the n=20 pull test, a lift-off of 3 or more times was considered a failure ("X"), 2 times was considered a "B" as there was room for improvement but it was unlikely to be a practical problem, 1 time was considered good ("A"), and 0 times was considered excellent ("S"). The results are shown in Table 1.

[0104] (Method for evaluating tool incompatibility) For each sample of copper wire with a diameter of 500 μm, an ultrasonic bonding device (K&S wire bonder ASTERION) was used to bond the wire to a copper plate so that the distance between the first and second joints was 5 mm. The second joint was bonded by bending the wire horizontally at a target angle of 45° relative to the wire axis. The bonding conditions were set to the optimal ultrasonic energy and pressure for each sample. The bonding tool used was Kulicke & Soffa, model number 127595-20, with wire-gripping alligator clips measuring 0.5 mm in width (inner diameter), 0.2 mm in depth (height), and 1.0 mm in length (length).

[0105] The determination of whether a tool detachment malfunction occurred was made by observing the condition of the wire at the second joint. For each sample, the combination of the first and second joints was performed 100 times, counting it as one attempt. If the wire did not adhere or there were four or more contact marks with the tool that made uneven contact, it was marked as a failure ("X"). If there were two or three contact marks, it was marked as "B" as some improvement is desired but there is no practical problem. If there was one contact mark, it was marked as good ("A") and the absence of any contact marks was marked as excellent ("S"), and this was used to evaluate the tool detachment. The evaluation results are shown in Table 1.

[0106] Furthermore, copper wires for Examples 31-111 and Comparative Examples 11-30 were obtained in the same manner as in Example 1 and Comparative Example 2, except for changes in composition and manufacturing conditions, and were evaluated in the same manner as in Example 1. The results are shown in Tables 4-6.

[0107] (comprehensive evaluation) In Table 1, 4-6 above, if there is one or more "S" ratings and the others are either "S" or "A", the overall rating is "Excellent", meaning it is outstanding. If there are two or more "A" and "S" ratings combined, the overall rating is "Good", meaning it is satisfactory. If there are two or more "B" ratings and no "X" ratings, the overall rating is "Acceptable", meaning it is a passing grade. If there is even one "X" rating, the overall rating is "Fail", meaning it is unacceptable. These ratings are recorded in Table 1. The specific combinations of each evaluation (the order does not matter) are as follows: "Excellent": SSS, SSA, SAA "Good": SAB, SSB, AAB, AAA "OK": SBB, ABB, BBB, "Not allowed": If there is at least one X.

[0108] [Table 1]

[0109] Table 2 shows the copper purity and some of the elemental concentrations (percentages) for the examples and comparative examples. Each concentration was calculated using an inductively coupled plasma mass spectrometer (ICP-MS) owned by the applicant, Tanaka Electronics Industry Co., Ltd. (Agilent Technologies, Inc. 8800 ICP-MS).

[0110] Next, the samples from the examples and comparative examples were left at room temperature (35°C) and humidity (75% RH) for 6 months. Afterward, the oxidation state of the copper wire surface was analyzed. The results showed that samples with an Ag concentration of 5 ppm by mass or higher had a copper(II) oxide (i.e., CuO) oxide film proportion of 20% or less. During the oxidation process of copper, copper(I) oxide (i.e., Cu2O) is usually formed first, and as oxidation progresses, the proportion of CuO increases. Therefore, by calculating the proportion of the CuO oxide film, the ease of oxidation formation (oxidation resistance) of the copper wire can be evaluated. The proportion of the CuO oxide film was calculated by converting "(CuO oxide film) / (CuO oxide film + Cu2O oxide film)" to a percentage.

[0111] The oxide film thickness was measured by Sequential Electrochemical Reduction Analysis (SERA), and the thickness equivalent value determined by SERA was used. The SERA analysis to determine the thickness of copper(II) oxide (CuO) and copper(I) oxide (Cu2O) mentioned above can be performed using an ECI Technology QC-200, for example, by the following procedure: A 0.5 mm diameter wire is sandwiched between two O-rings, a 2.1 cm diameter region is isolated with a gasket, borate buffer is injected, and the region is saturated with nitrogen. A 150 μA / cm² current is applied to the region. 2 A current density (I) is applied, and the time (seconds) (t) for the Cu2O reduction reaction occurring between -0.30V and -0.60V, and the CuO reduction reaction occurring between -0.60V and -0.85V is measured. The thickness T (nm) of CuO and Cu2O is calculated using the constant K obtained from Faraday's law, based on T = K·I·t. The value of the constant K for CuO is 6.53 × 10⁻⁶. -5 (cm 3 The formula is (A·sec), and the value of K for Cu2O is 2.45 × 10⁻⁶. -4 (cm 3 It is ( / A·sec).

[0112] Measurements were taken at three locations on the wire: the area around the tip, the area around the trailing end, and the intermediate section. The average value of these measurements was then calculated. Samples with a CuO oxide film content of 20% or less were rated "S" (indicating excellent performance), while samples with a content between 20% and 30% were rated "A" (indicating good performance).

[0113] Here, we will explain why a CuO oxide film ratio of 30% or less was deemed good. In power modules, IGBTs, MOSFETs, and diode chips are mounted on a substrate where a ceramic plate (such as silicon nitride or alumina) and a copper plate are bonded together using either Active Metal Brazing (AMB) or Direct Copper Bonding (DCB), and circuits are formed on the substrate by wire bonding.

[0114] For example, in copper wires bonded to a copper plate on a DCB substrate, a problem can occur where the wire delaminates due to stress on the wire joint caused by thermal expansion and contraction of the silicone resin during a temperature cycle test (e.g., -65°C to 150°C) performed on the finished product with the silicone resin sealed inside. The temperature conditions for this temperature cycle test are determined by a globally standardized automotive specification. The inventors considered that the growth of the oxide film on the copper wire surface due to heat is one of the main causes of this delamination. They found that if the proportion of the CuO film on the copper wire surface after a standing test, i.e., after 6 months of standing at room temperature of 35°C and humidity of 75%RH, is 30% or less, the delamination of the wire joint can be significantly reduced in the aforementioned temperature cycle test.

[0115] Furthermore, in recent years, the operating temperature range for power semiconductors has broadened, and there are cases where guarantees are required in a range of -65°C to 175°C, for example. In these temperature cycling tests, a higher load is placed on the copper wire compared to the test conditions described above, making wire delamination from the joint more likely. The inventors of the present invention discovered that when the ratio of the CuO oxide film on the wire surface can be reduced to 20% or less in the aforementioned standing test, these high requirements can be met, and wire delamination from the joint does not occur even under more severe conditions, resulting in stable joint strength.

[0116] Furthermore, in addition to the problem of wire delamination due to the formation of a copper oxide film, the inventors have found that expanding the above-mentioned guaranteed temperature range can also lead to a problem of decreased electrical conductivity at high temperatures. In other words, electrical conductivity tends to decrease as the temperature increases. As mentioned above, the diameter of the copper wire in the embodiment is thicker than that of a typical semiconductor bonding wire, which has the advantage of being able to carry a large current at a high voltage. However, as the temperature rises, the electrical resistance increases and the electrical conductivity decreases, which may cause the current value to decrease and the driving force to decrease.

[0117] Therefore, in the manufacturing of power semiconductor devices, it is necessary to quantitatively predict the increase in electrical resistance of copper bonding wires at higher temperatures. If the electrical resistance increases beyond this prediction, it may not only fail to achieve the current value specified in the design, but the heat generated by the wire due to the increased electrical resistance may also adversely affect surrounding components. For example, there is a risk of serious problems occurring, such as the sealing resin, a surrounding component, melting due to the heat generated by the wire.

[0118] The inventors have diligently conducted repeated research and experiments to expand the guaranteed temperature range and, in particular, to investigate the rate of increase in electrical resistance of copper bonding wires at high temperatures and the resulting heat generation. As a result, they discovered that if the electrical conductivity (IACS) at room temperature is 99% or higher (preferably 99.5% or higher), the increase in electrical resistance falls within the range normally expected in the manufacture of power semiconductor devices (within approximately 1.7 times).

[0119] Furthermore, in order to suppress the increase in electrical resistance and keep the proportion of CuO oxide film on the copper wire surface below 30%, the inventors repeatedly conducted experiments to evaluate the electrical resistance and CuO oxide film formation by adding trace amounts of various elements to high-purity copper. As a result, they found that both of these problems could be solved simultaneously by including silver (Ag) at a concentration of 5 to 30 ppm by mass relative to the entire copper wire, and phosphorus (P), iron (Fe), silicon (Si), arsenic (As), and antimony (Sb) at a concentration of 0 to 3 ppm by mass, with the total amount of these elements other than silver being 15 ppm or less by mass.

[0120] Electrical conductivity was measured using the IACS (International Annealed Copper Standard) by taking measurements of the same area and calculating the average value. The IACS method used a four-terminal approach with a potential lead distance of 100 mm and a DC current of 200 mA to determine the electrical resistivity of each wire. The electrical resistivity of annealed standard soft copper at room temperature (20°C) was 1.7241 × 10⁻¹⁰. -2This value is expressed as a ratio with μΩm set as 100% of IACS. Compared to IACS of 99.999 mass% pure copper, samples with an IACS decrease rate of less than 0.5% are rated "S" to indicate excellent performance, and samples with a decrease rate between 0.5% and 1% are rated "A" to indicate good performance, as shown in Table 2. In the comparative examples, some also received evaluations of "S" or "A," but these evaluations were specific to the proportion of CuO oxide film and the decrease rate of IACS. As shown in Table 1, other evaluations were "X," so the overall evaluation was judged as "Unacceptable."

[0121] Furthermore, the copper purity and trace element content were measured in the same manner as described above for the copper wires of Examples 31-111 and Comparative Examples 11-30 shown in Tables 4-6, and the proportion of CuO oxide film (CuO ratio) and the IACS reduction rate were evaluated. The results are shown in Tables 7-9.

[0122] [Table 2]

[0123] Using the same samples as the examples and comparative examples shown in Table 1, we evaluated whether similar cutting performance could be obtained when using the rear-cut method (also called the back-cut method or half-cut method). The evaluation was performed using the same method as described in the "Cutting Performance Evaluation Method" above, except that the wire was not cut completely (cut to a depth of approximately 50-70% of the diameter of the cross-section perpendicular to the longitudinal direction of the wire). The results are shown in Table 3.

[0124] [Table 3]

[0125] As shown in Table 3, the cutting performance of the wire samples obtained using the rear-cut method was generally worse than that of the commonly used full-cut method, but still achieved an acceptable level of performance. This confirms that the copper wire used in the example can be used without problems even with the rear-cut method. It should be noted that some of the comparative examples also received an "A" rating, but this was based solely on the cutting performance. As shown in Table 1, it was confirmed that other evaluation items resulted in an "X" rating, leading to an overall "Unacceptable" rating.

[0126] This section explains the front-cut method (also called the full-cut method) and the rear-cut method. Figure 11 is a schematic diagram showing the structure of the bond head 111 and its movement after the second bond in the front-cut method. The bond head 111 is part of the bonder and is equipped with a bonding tool and cutter that grips the wire. When the first and second bonds are performed sequentially, the bond head 111 moves the gripped wire from the first bond to the second bond and bonds them to the respective targets. In a front-cut bonder, the cutter is located in front of the bond head. Therefore, after the second bond, the bond head (cutter) is moved to the planned wire cutting position. That is, the bond head is moved in the direction indicated by the arrow in Figure 11, and the wire is cut at that position. The distance from the position at the time of the second bond to the wire cutting position is approximately 3 mm, although this depends on the type of bond head 111. Therefore, as shown in Figure 11, if there is an obstacle such as the semiconductor device case 114 in the direction of movement of the bond head 111 (arrow direction), for example, if the distance from the bonding point to the obstacle is 3 mm or less, the bond head 111 may collide with the case 114, causing the bonder to stop or the semiconductor device to be damaged.

[0127] Figure 12 is a schematic diagram showing the structure of the bond head in the rear-cut method and the state immediately after wire bonding. The movement of the bond head 111 during the formation process of the first and second bonds in the rear-cut method is the same as in the front-cut method, but since the position of the cutter 112 during the second bond is almost the same as the position where the wire 113 is cut, there is no need to move the bond head (cutter) when cutting the wire. Therefore, even if an obstacle such as the semiconductor device case 114 is located close to the bonding site, the possibility of the bond head coming into contact with the obstacle is much smaller than in the front-cut method. As semiconductor devices continue to become smaller in the future, expectations for the rear-cut bonding method are increasing.

[0128] Furthermore, when performing reverse bonding (where, in normal bonding, the first bond is to the IC chip and the second bond is to the external electrodes or circuit pattern on the substrate, but in reverse bonding, the copper wire is connected to the external electrodes or circuit pattern as the first bond, and the copper wire is bonded to the IC chip as the second bond), the second bond immediately before wire cutting is on the IC chip. In the front-cut method, i.e., the full-cut method, there is a risk that the cutter blade will come into contact with the particularly fragile IC chip, damaging it. Therefore, in reverse bonding, the rear-cut method, i.e., the half-cut method, which does not cut the wire 100%, but makes a cut about 50% of the wire and then tears it off, is suitable. Accordingly, it was confirmed that the bonding wire of this embodiment, which has excellent cutting properties, exhibits its cutting effect not only in the front-cut method but also in the rear-cut method.

[0129] [Table 4]

[0130] [Table 5]

[0131] [Table 6]

[0132] [Table 7]

[0133] [Table 8]

[0134] [Table 9]

[0135] Next, we will describe the case where the copper wire of the embodiment is applied to the connection between the busbar of a lithium-ion battery module in an electric vehicle and the electrodes of the lithium-ion battery. A lithium-ion battery consists of dozens of columnar lithium-ion batteries as a single unit, and a gel-like substance is filled in the gaps between adjacent lithium-ion batteries to prevent contact between adjacent batteries and to absorb vibrations during vehicle operation.

[0136] Figure 13 is a schematic diagram showing a part of a lithium-ion battery module. The lithium-ion battery module shown in Figure 13 comprises a lithium-ion battery 51, a busbar 52 on the cathode side, a busbar 53 on the anode side, and copper wires 50 that electrically connect the lithium-ion battery 51 and the busbar 52 on the cathode side. The lithium-ion battery module shown in Figure 13 also comprises another copper wire 50 that electrically connects the lithium-ion battery 51 and the busbar 53 on the anode side. The copper wires 50 are the same as those used in the embodiments described above. In the lithium-ion battery module shown in Figure 13, the anode and cathode are arranged on the top and bottom surfaces of the columnar lithium-ion battery, respectively, and it is usually mounted in a vehicle with the anode facing up and the cathode facing down.

[0137] Figure 14 is a schematic diagram showing a part of another form of lithium-ion battery module. The lithium-ion battery module shown in Figure 14 comprises a lithium-ion battery 51, a cathode busbar 52, an anode busbar 53, and copper wires 50 that electrically connect the lithium-ion battery 51 and the cathode busbar 52. The lithium-ion battery module shown in Figure 14 also further comprises other copper wires 50 that electrically connect the lithium-ion battery 51 and the anode busbar 53. The copper wires 50 are the same as those used in the embodiments described above. In the lithium-ion battery module shown in Figure 14, both the anode and cathode are positioned at the center and outer edge of the top surface of the columnar lithium-ion battery, respectively, and it is usually mounted in a vehicle with the anode facing upwards.

[0138] Figure 15 is a schematic diagram of the entire lithium-ion battery module. As shown in Figure 15, the lithium-ion battery module is constructed by housing multiple lithium-ion batteries 51 and filling the spaces between the multiple lithium-ion batteries 51 with a gel-like substance 54. Copper wires 50 are connected to the lithium-ion batteries 51.

[0139] In the joining of the busbar and copper wire in this lithium-ion battery module, and the joining of the lithium-ion battery electrodes and copper wire (wire bonding), good bonding strength can be maintained by using copper wire with good conformability. The reasons for this are as follows: Because ultrasound is used in wire bonding, and because the lithium-ion battery is not fixed in a gel-like material, the lithium-ion battery itself vibrates due to the ultrasound during wire bonding. In addition, there is a height difference between the lithium-ion battery electrodes and the busbar. Therefore, in wire bonding, bending operations are performed that take into account the vibration of the lithium-ion battery and the height difference between the lithium-ion battery and the busbar. By using copper wire with a high degree of freedom in such bending, i.e., high conformability, problems such as the wire coming off the tool during bonding and insufficient bonding strength can be suppressed, and breakage due to vibration of the copper wire while the vehicle is in motion can be prevented.

[0140] As miniaturization of batteries is required for the miniaturization of electric vehicles, the copper wire of the embodiment with excellent conformability can achieve excellent bonding strength even with complex bending operations in confined spaces. Furthermore, the copper wire of the embodiment with low electrical resistance (high electrical conductivity) does not generate much heat, which can enhance the safety of lithium-ion batteries that are at risk of explosion at high temperatures (e.g., above 80°C).

[0141] (Evaluation of the bonding between the battery electrodes and the busbar, etc.) Using the copper wires from Examples 1-30 and Comparative Examples 1-3, two copper plates were used to match the actual bonding conditions between lithium-ion battery electrodes and busbars. The first bond was made on the first plate, and the second bond was made on the second plate, which was positioned approximately 20 mm lower than the first plate. The two plates used for the experimental evaluation were not only positioned at different heights, but also at an angle of approximately 45 degrees when the first and second bonds were projected onto a plane. Except for the difference in the positional relationship between the first and second bonds, the cutting performance, bonding strength, and tool release were evaluated under the same conditions as in the above-mentioned examples using the same HESSE BJ935 bonder. As a result, all samples from Examples 1-30 received a rating of "B" or higher. The copper wire in Comparative Example 1 received an "X" rating for cutting performance, the copper wire in Comparative Example 2 received an "X" rating for bonding strength, and the copper wire in Comparative Example 3 received an "X" rating for tool release.

[0142] Based on the above, the copper wire of the embodiment, in wire bonding inside power semiconductor devices, etc., was able to simultaneously solve the problems of good cutting performance, wire peeling (lift-off) from the wedge joint, and wire detachment from the wedge tool due to improved wire followability during complex bonding operations involving sharp lateral bending, by controlling its crystal grain boundary density, crystal orientation ratio, dynamic hardness, and elastic modulus within a predetermined range.

[0143] Furthermore, the copper wire of this embodiment exhibits excellent conformability, a low CuO oxide film formation ratio, high bonding strength, and high electrical conductivity, making it highly suitable not only for power semiconductor applications but also for bonding busbars and electrodes of lithium-ion batteries in lithium-ion battery modules for electric vehicles.

[0144] The copper wire of this embodiment can greatly contribute to the development of industries such as the automotive industry, power electronics industry, electric railway industry, and power industry, and in turn, greatly contribute to reducing greenhouse gas emissions and preventing global warming.

Claims

1. A copper bonding wire made of a copper alloy with a copper purity of 99.99% by mass or higher, The grain boundary density in the cross-section perpendicular to the wire axis of the copper bonding wire is 0.01 μm. －１ The above 0.6 μm －１ It is less than, In the aforementioned cross-section, among the crystal orientations in the wire axis direction, the ratio of crystal orientation <111>, where the angular difference with respect to the wire axis direction is 15° or less, is divided by the ratio of crystal orientation <101>, and the value obtained is 10 or more and 650 or less. The dynamic hardness in the cross-section is 45 or more and 90 or less. Furthermore, the copper bonding wire is characterized in that the elastic modulus in the cross-section is 20 GPa or more and 70 GPa or less.

2. The copper bonding wire according to claim 1, wherein the copper alloy contains 5 ppm to 30 ppm of silver (Ag) based on its total weight, and contains 0 ppm to 3 ppm of phosphorus (P), iron (Fe), silicon (Si), arsenic (As), and antimony (Sb) each, totaling 15 ppm or less.

3. A copper bonding wire according to claim 1 or claim 2, wherein the wire diameter is 40 μm or more and 700 μm or less.

4. A method for producing a copper bonding wire made of a copper alloy with a copper purity of 99.99% by mass or higher, A process of preparing copper wires made of a copper alloy with a mass of 99.99% or more, The process of drawing the aforementioned copper wire, The process includes a step of applying a final heat treatment to the drawn copper wire, The grain boundary density in the cross-section perpendicular to the wire axis of the copper bonding wire is 0.01 μm. －１ The above 0.6 μm －１ It is less than, In the aforementioned cross-section, among the crystal orientations in the wire axis direction, the ratio of crystal orientation <111>, where the angular difference with respect to the wire axis direction is 15° or less, is divided by the ratio of crystal orientation <101>, and the value obtained is 10 or more and 650 or less. The dynamic hardness in the cross-section is 45 or more and 90 or less. A method for manufacturing a copper bonding wire having an elastic modulus of 20 GPa or more and 70 GPa or less in the cross-section.

5. At least one substrate, at least one semiconductor element on the substrate, an electrode on the surface of the semiconductor element, a semiconductor element substrate on the semiconductor element, and at least one circuit pattern on the substrate. A power semiconductor device comprising electrodes on the surface of the semiconductor element, electrodes on the surface of the semiconductor element and an external electrode, electrodes on the surface of the semiconductor element and one of the circuit patterns, electrodes on the surface of the semiconductor element and a terminal, two adjacent circuit patterns from the circuit patterns, one of the circuit patterns and a terminal, substrates and substrates, and copper wires connecting one or more selected from the group consisting of the semiconductor element substrate and the circuit patterns, The aforementioned copper wire is A copper bonding wire made of a copper alloy with a copper purity of 99.99% by mass or higher, The grain boundary density in the cross-section perpendicular to the wire axis of the copper bonding wire is 0.01 μm. －１ The above 0.6 μm －１ It is less than, In the aforementioned cross-section, among the crystal orientations in the wire axis direction, the ratio of crystal orientation <111>, where the angular difference with respect to the wire axis direction is 15° or less, is divided by the ratio of crystal orientation <101>, and the value obtained is 10 or more and 650 or less. The dynamic hardness in the cross-section is 45 or more and 90 or less. Furthermore, the elastic modulus of the copper bonding wire in the cross-section is 20 GPa or more and 70 GPa or less. Semiconductor equipment.

Images