Al bonding wire or Al bonding ribbon

Abstract

To provide an Al bonding wire or Al bonding ribbon that exhibits excellent high-speed temperature cycle reliability even in high-speed temperature cycle tests with high upper limit temperatures, which are required for next-generation SiC power semiconductors. The Al bonding wire or Al bonding ribbon contains 3.0% by mass or more and 20.0% by mass or less of Si, and S0 is a closed curve representing the periphery of a cross section perpendicular to the central axis of the Al bonding wire or Al bonding ribbon, S1 is a closed curve that is 2 / 3 times similar to S0, and S2 is a closed curve that is 1 / 3 times similar to S0, and S0, S1, and S2 are arranged so that their centers of gravity coincide with the center of gravity of the cross section perpendicular to the central axis of the Al bonding wire or Al bonding ribbon, and the area surrounded by S0 and S1 is the surface portion, and the area surrounded by S2 is the core portion, and when the volume fraction of voids in the core portion having a sphere-equivalent diameter of 1 μm or more and less than 10 μm is Rd, and the volume fraction of voids in the surface portion having a sphere-equivalent diameter of 1 μm or more and less than 10 μm is Rf, the ratio (Rf / Rd) is 0.005 or more and 0.50 or less.

Description

[Technical Field]

[0001] The present invention relates to an Al bonding wire or an Al bonding ribbon, and further to a semiconductor device obtained using the Al bonding wire or the Al bonding ribbon. [Background technology]

[0002] In semiconductor devices, electrodes formed on a semiconductor chip are connected to electrodes on a lead frame or substrate using bonding wires (wire material) or bonding ribbons (strip material). Power semiconductor devices mainly use bonding wires or bonding ribbons made of aluminum (Al). The wire diameter of Al bonding wires is mainly in the range of 100 μm to 600 μm, while the width of Al bonding ribbons is mainly in the range of 100 μm to 3000 μm and the thickness is mainly in the range of 50 μm to 600 μm. Here, Al bonding wires and Al bonding ribbons are collectively referred to as Al connecting materials.

[0003] In power semiconductor devices, silicon (Si) is often used as the material for the semiconductor chip, and Al-Si alloys or Al-Cu alloys are often used as the materials for the electrodes formed on the semiconductor chip. Power semiconductor devices using Al bonding wire or Al bonding ribbon are often used in high-power equipment such as air conditioners and solar power generation systems, as well as in-vehicle semiconductor devices.

[0004] There are two methods for joining Al bonding wire or Al bonding ribbon: the first bonding with an electrode on a semiconductor chip, and the second bonding with an electrode on a lead frame or substrate, both of which use wedge joining. Wedge joining is a method in which ultrasonic vibration and load are applied to the Al bonding wire or Al bonding ribbon via a metal jig (tool), destroying the surface oxide film between the Al bonding wire or Al bonding ribbon and the electrode material, exposing a new surface and performing solid-state diffusion bonding. This joining method is characterized by connecting in a solid state without melting the connecting material, and is different from welding techniques, which melt the connecting material.

[0005] Next-generation power semiconductor devices are required to operate stably for longer periods of time than general-purpose power semiconductor devices. Power semiconductor devices operate by repeatedly turning current on and off. When current is supplied to a Si semiconductor chip through an Al bonding wire or Al bonding ribbon, the temperature of the first junction rises. On the other hand, when the current supply is stopped, the temperature of the first junction drops. Thus, the first junction repeatedly rises and falls in temperature during power semiconductor operation. This repeatedly applies thermal stress to the first junction due to the difference in thermal expansion between the Al bonding wire or Al bonding ribbon and the semiconductor chip. When a connecting material made solely of high-purity Al is used, the Al bonding wire or Al bonding ribbon breaks down due to thermal stress in a relatively short period of time, making it difficult to achieve the performance required for next-generation power semiconductor devices. Therefore, next-generation power semiconductors are required to improve the junction life (hereinafter also referred to as "temperature cycle reliability") associated with temperature rise and fall of the first junction.

[0006] In response to the demand for temperature cycle reliability, Al bonding wires have been proposed that focus on improving mechanical strength. Adding specific elements to Al has been proposed as a method for improving the mechanical properties of Al bonding wires.

[0007] Patent Document 1 discloses a bonding wire made of an Al alloy containing at least magnesium (Mg) and silicon (Si), with the total content of Mg and Si being 0.03% by mass or more and 0.3% by mass or less. This patent document discloses that the decrease in bonding strength of the first bonded portion in a cold-temperature cycle test in the temperature range of 70°C to 120°C is delayed due to the effect of increasing strength through solid solution strengthening of Mg and Si and the effect of suppressing crack propagation due to precipitated magnesium silicide (MgSi).

[0008] Patent Document 2 discloses a bonding wire made of an alloy containing 0.01 to 0.2 mass% iron (Fe), 1 to 20 mass ppm silicon (Si), and the remainder being Al with a purity of 99.997 mass% or more, wherein the amount of Fe in solid solution is 0.01 to 0.06%, the amount of Fe precipitated is 7 times or less the amount of Fe in solid solution, and the wire has a fine structure with an average crystal grain size of 6 to 12 μm. This patent document also discloses that by uniformly dispersing intermetallic compound particles of Fe and Al in Al to improve the mechanical strength of the matrix and further refining the recrystallized grains, it is possible to suppress a decrease in the bonding strength of the first bonded portion in a thermal shock test in a temperature range of -50°C to 200°C.

[0009] Patent Document 3 discloses a bonding wire obtained by melting an Al-Si alloy containing 0.1 to 5 mass % silicon (Si) with the remainder being Al and impurities, and then forming the melted Al-Si alloy into a thin wire by rapid cooling. This patent document discloses that the mechanical strength is improved by rapidly cooling the molten Al-Si alloy to finely and uniformly disperse the Si. [Prior art documents] [Patent documents]

[0010] [Patent Document 1] Japanese Patent Application Laid-Open No. 2014-131010 [Patent Document 2] Japanese Patent Application Laid-Open No. 2014-129578 [Patent Document 3] Japanese Patent Application Publication No. 59-57440 Summary of the Invention [Problem to be solved by the invention]

[0011] As mentioned above, next-generation power semiconductor devices are required to withstand longer periods of use than general-purpose power semiconductor devices. During operation of a power semiconductor device, the temperature of the first bonded portion repeatedly rises and falls. As a result, because the Al bonding wire or Al bonding ribbon has a larger linear expansion coefficient than the semiconductor chip, thermal stress occurs at the first bonded portion due to the difference in linear expansion coefficients between the two, which can ultimately lead to fatigue failure of the Al bonding wire or Al bonding ribbon. A temperature cycle test is one type of accelerated evaluation test for the lifespan (temperature cycle reliability) of such a first bonded portion as it rises and falls in temperature. The Al bonding wire or Al bonding ribbon used in next-generation power semiconductors is required to exhibit excellent temperature cycle reliability in a temperature cycle test.

[0012] However, the inventors have confirmed that when using an Al bonding wire that has been strengthened by adding Si or the like as disclosed in Patent Documents 1 to 3, in temperature cycle tests intended for use in next-generation power semiconductor devices, cracks propagate at a relatively fast rate within the Al alloy electrode, which has lower strength than the Al bonding wire, making it difficult to consistently achieve good temperature cycle reliability.

[0013] On the other hand, conventional temperature cycle tests (hereinafter also referred to as "TCT (Temperature Cycle Test)") can be easily performed using commercially available test equipment. However, the temperature change rate in TCT is relatively slow, which raises concerns about discrepancies with the fast temperature change rate during operation of power semiconductor devices. Therefore, recently, high-speed temperature cycle tests (hereinafter also referred to as "high-speed TCT"), which increase the temperature change rate to more closely resemble actual usage conditions, have been considered. The temperature change rate in conventional TCT is, for example, about 10°C / min, while in high-speed TCT, the temperature changes at a high rate of, for example, about 200°C / min. Regarding reliability evaluation of Al bonding wire or Al bonding ribbon joints, the inventors have confirmed that even Al bonding wires or Al bonding ribbons that do not show a decrease in reliability when evaluated using conventional TCT may experience a decrease in bond strength and a shortened bond life when evaluated using high-speed TCT. Therefore, there is a demand for Al bonding wire or Al bonding ribbon that exhibits good joint reliability even in high-speed TCT, a more severe test that is closer to the conditions of actual use, and that provides excellent temperature cycle reliability. Hereinafter, the temperature cycle reliability in high-speed TCT may be referred to as "high-speed temperature cycle reliability."

[0014] Furthermore, the use of silicon carbide (SiC), with its high heat resistance, is expected to increase in next-generation power semiconductor elements, replacing the previously mainstream silicon (Si). Connections for SiC power semiconductors will require even more stringent high-speed temperature cycle testing than currently available. For example, while the temperature range for high-speed temperature cycle testing for Si semiconductors is -40°C to 150°C, SiC semiconductors require high-speed temperature cycle reliability under the stricter temperature range of -40°C to 175°C. Furthermore, next-generation SiC power semiconductors, which will be used at high power output due to the SiC's heat resistance, will be required to exhibit excellent high-speed temperature cycle reliability even under more stringent test conditions, with the upper limit temperature rising to 185°C. However, if the upper limit temperature for high-speed TCT increases from 175°C to 185°C, the temperature difference in the thermal cycle increases by 10°C, magnifying the thermal expansion difference at the joint of the Al bonding wire or Al bonding ribbon, accelerating fatigue failure. Furthermore, since high temperatures are reached, the bonding condition of the bonding interface prior to the test can promote fatigue fracture, making it necessary to improve the initial bonding condition.

[0015] The present invention has been made in consideration of the above-mentioned problems, and aims to provide an Al bonding wire or Al bonding ribbon that exhibits excellent high-speed temperature cycle reliability even in high-speed temperature cycle tests with high upper limit temperatures, which are required for next-generation SiC power semiconductors. [Means for solving the problem]

[0016] As a result of intensive research into the above-mentioned problems, the inventors have discovered that an Al bonding wire or Al bonding ribbon containing 3.0 mass% or more and 20.0 mass% or less of Si can solve the above-mentioned problems if the ratio (Rf / Rd) is within a specific range, where Rd is the volume fraction of voids in the core having a sphere-equivalent diameter of 1 μm or more and less than 10 μm, and Rf is the volume fraction of voids in the surface having a sphere-equivalent diameter of 1 μm or more and less than 10 μm, as measured by X-ray CT (Computed Topography) analysis.Based on this finding, the inventors have conducted further research and have completed the present invention.

[0017] That is, the present invention includes the following. <1> An Al bonding wire or Al bonding ribbon containing 3.0 mass% or more and 20.0 mass% or less of Si, The closed curve representing the periphery of the cross section perpendicular to the central axis of the Al bonding wire or Al bonding ribbon is S0, the closed curve which is 2 / 3 times similar to S0 is S1, and the closed curve which is 1 / 3 times similar to S0 is S2. S0, S1, and S2 are arranged so that their centers of gravity coincide with the center of gravity of the cross section perpendicular to the central axis of the Al bonding wire or Al bonding ribbon, and the area surrounded by S0 and S1 is the surface portion, and the area surrounded by S2 is the core portion. An Al bonding wire or Al bonding ribbon in which the ratio (Rf / Rd) is 0.005 or more and 0.50 or less, when Rd is the volume fraction of voids in the core portion with a sphere-equivalent diameter of 1 μm or more and less than 10 μm, and Rf is the volume fraction of voids in the surface portion with a sphere-equivalent diameter of 1 μm or more and less than 10 μm, as measured by X-ray CT (Computed Topography) analysis. <2> When the number of Si phases having a circle equivalent diameter of 0.5 μm or more and 0.8 μm or less in the L cross section (a cross section in the central axis direction including the central axis) is Ns, and the number of Si phases having a circle equivalent diameter of 0.5 μm or more in the L cross section is Nc, the ratio of Ns to Nc [Ns / Nc × 100 (%)] is 30% or more and 95% or less, <1> The Al bonding wire or Al bonding ribbon according to claim 1. <3> Furthermore, it contains one or more of Sr, Na, Eu, and P in a total amount of 10 mass ppm or more and 800 mass ppm or less. <1> or <2> The Al bonding wire or Al bonding ribbon according to claim 1. <4> Further, it contains one or more of Ni, Ti, Fe, and Mg in a total amount of 100 mass ppm or more and 1500 mass ppm or less. <1> ~ <3> The Al bonding wire or Al bonding ribbon according to any one of the above. <5> The total concentration of elements other than Al, Si, Sr, Na, Eu, P, Ni, Ti, Fe, and Mg in the Al bonding wire or Al bonding ribbon is 0.5 mass% or less. <1> ~ <4> The Al bonding wire or Al bonding ribbon according to any one of the above. <6> The circle equivalent diameter and the number of Si phases are values ​​measured using a SEM-EDS-EBSD device. <2> ~ <5> The Al bonding wire or Al bonding ribbon according to any one of the above. <7> for semiconductor device, <1> ~ <6> The Al bonding wire or Al bonding ribbon according to any one of the above. <8> <1> ~ <7> A semiconductor device comprising the Al bonding wire or Al bonding ribbon according to any one of the above. [Effects of the Invention]

[0018] According to the present invention, it is possible to provide an Al bonding wire or Al bonding ribbon that exhibits excellent high-speed temperature cycle reliability even in high-speed temperature cycle tests with high upper limit temperatures, which are required for next-generation SiC power semiconductors, and a semiconductor device obtained using the Al bonding wire or Al bonding ribbon. [Brief explanation of the drawings]

[0019] [Figure 1] 1 is a schematic diagram illustrating a measurement surface (inspection surface) when measuring the small diameter ratio of the Si phase in an Al bonding wire. The measurement surface is a cross section (L cross section) in the central axis direction including the central axis of the Al bonding wire. [Figure 2] 2 is a schematic diagram illustrating the measurement surface (inspection surface) when measuring the small diameter ratio of the Si phase in an Al bonding ribbon. The measurement surface is a cross section (L cross section) in the central axis direction including the central axis of the Al bonding ribbon. [Figure 3] FIG. 3 shows an example of a brightness histogram created when determining the brightness thresholds for voids or external spaces and materials in X-ray CT analysis. [Figure 4] FIG. 4 is an example of a graph showing the number distribution of the circle-equivalent diameter of the Si phase in the L cross section. [Figure 5] FIG. 5 is a schematic diagram showing the relationship between S0, S1, and S2, as well as the core, intermediate, and surface portions, in the case of an Al bonding wire. [Figure 6] FIG. 6 is a schematic diagram showing the relationship between S0, S1, and S2, and the core, middle, and surface regions in the case of an Al bonding ribbon. DETAILED DESCRIPTION OF THE INVENTION

[0020] The present invention will be described in detail below with reference to preferred embodiments. While the description may refer to drawings, each drawing merely shows the shape, size, and arrangement of components to the extent that the invention can be understood. The present invention is not limited to the following embodiments and examples, and can be modified and implemented as desired within the scope of the claims of the present invention and their equivalents.

[0021] [Al bonding wire or Al bonding ribbon] The Al bonding wire or Al bonding ribbon of the present invention is an Al bonding wire or Al bonding ribbon containing 3.0 mass% or more and 20.0 mass% or less of Si, The closed curve representing the periphery of the cross section perpendicular to the central axis of the Al bonding wire or Al bonding ribbon is S0, the closed curve which is 2 / 3 times similar to S0 is S1, and the closed curve which is 1 / 3 times similar to S0 is S2. S0, S1, and S2 are arranged so that their centers of gravity coincide with the center of gravity of the cross section perpendicular to the central axis of the Al bonding wire or Al bonding ribbon, and the area surrounded by S0 and S1 is the surface portion, and the area surrounded by S2 is the core portion. The ratio (Rf / Rd) (hereinafter also referred to as "void gradient ratio") is 0.005 or more and 0.50 or less, where Rd is the volume fraction of voids in the core portion having a sphere-equivalent diameter of 1 μm or more and less than 10 μm, and Rf is the volume fraction of voids in the surface portion having a sphere-equivalent diameter of 1 μm or more and less than 10 μm, as measured by X-ray CT (Computed Topography) analysis.

[0022] The Al bonding wire or Al bonding ribbon of the present invention contains 3.0 mass% or more and 20.0 mass% or less of Si, and has an Al phase in which Si is solid-solved in Al, and a Si phase formed by crystallization or precipitation of Si. Here, the Al phase may contain other additive elements in addition to Si as a solid solution. The Si phase is a general term for Si crystallized deposits and Si precipitates. Si crystallized deposits are formed from the melt during solidification and are coarse, measuring approximately 1 to 25 μm in size, while Si precipitates are formed from the solid state and are small, measuring approximately 0.1 to several μm in size. The Si phase has a smaller linear expansion coefficient than Al, which contributes to reducing the difference in linear expansion coefficient between the Al bonding wire or Al bonding ribbon and the semiconductor chip, thereby reducing thermal stress and improving high-speed temperature cycle reliability.

[0023] In the present invention, "wire" and "ribbon" are not classified by shape, but by their manufacturing method. That is, "wire" refers to a "connecting material manufactured by wire drawing using a die," and "ribbon" refers to a "connecting material manufactured by a rolling process." A "wire" typically has a circular cross-sectional shape, and a "ribbon" typically has a rectangular or nearly rectangular cross-sectional shape.

[0024] That is, the central axis of an Al bonding wire and a cross section (L cross section) in the direction of the central axis including the central axis are as shown in FIG. 1. FIG. 1 shows an Al bonding wire having a circular cross section, but in the case of an Al bonding ribbon having a rectangular or approximately rectangular cross section with a width W and a thickness T, the central axis refers to the axis passing through the center of the width W and the center of the thickness T, and the L cross section refers to a cross section in the direction of the central axis including the central axis and perpendicular to the width W direction (FIG. 2). Specifically, the central axis of an Al bonding ribbon and a cross section (L cross section) in the direction of the central axis including the central axis are as shown in FIG. 2. When processing the cross section to expose the L cross section of the Al bonding wire, there may be a deviation from the central axis of the Al bonding wire. In this case, if the length of the L cross section in the direction perpendicular to the central axis is 90% or more of the wire diameter of the Al bonding wire, it can be considered to be a cross section including the central axis.

[0025] -Gap gradient ratio- As mentioned above, when using a connecting material consisting solely of high-purity Al, cracks propagate relatively rapidly within the wire or ribbon during temperature cycle testing, resulting in reduced temperature cycle reliability. It has been confirmed that Al alloys containing high concentrations of Si can improve temperature cycle reliability by reducing the thermal expansion of the wire or ribbon. It has also been confirmed that high-speed TCT, which is similar to actual operating conditions, can sometimes result in reduced bond strength and shortened bond life, even for Al bonding wires or ribbons that do not exhibit a decrease in reliability when evaluated using conventional TCT. This is thought to be due to the faster temperature change rate in high-speed TCT, which increases thermal stress at the bond compared to conventional TCT. Furthermore, when the upper limit temperature of high-speed TCT is higher, such as 185°C, bond strength is more likely to decrease. Therefore, further improvements in high-speed temperature cycle reliability at high temperatures are required to meet the high-speed temperature cycle reliability required for next-generation power semiconductor devices, such as SiC, which have high heat resistance.

[0026] Increasing the upper temperature limit of high-speed temperature cycling tests increases the difference in thermal expansion between the Al bonding wire / ribbon and the SiC semiconductor, resulting in increased thermal strain at the wire or ribbon joint. The formation of a Si phase in wires or ribbons made of Al alloys with a high Si content (hereinafter referred to as "high-concentration Al-Si alloys") reduces thermal expansion, which advantageously reduces the thermal strain of the entire wire or ribbon and suppresses cracking within the wire or ribbon (hereinafter referred to as "internal wire cracking"). However, when observing the failure mode during high-speed TCT with an upper temperature limit of 185°C, we found that the cause of the failure was cracking in the Al-based electrode film (hereinafter referred to as "internal electrode cracking"). Because the Al-based electrode film is softer than the high-concentration Al-Si alloy wire or ribbon, the thermal stress from high-temperature high-speed TCT propagates to the electrode, causing cracking within the electrode. The failure mechanism of high-speed TCT changes from the general crack in the wire to the crack in the electrode, so material design that focuses on improving the crack in the electrode is required.

[0027] As a result of intensive research to solve the above problems, the inventors have found that forming minute voids within a high-concentration Al-Si alloy wire or ribbon, and further having a gradient structure in which the distribution of void volume within the wire or ribbon changes from the surface region to the central region, contributes to suppressing cracks within the electrode during high-temperature, high-speed TCT tests.

[0028] Specifically, the inventors discovered that voids on the order of several micrometers exist in Al-Si alloy wires or ribbons, primarily around the Si phase. This is believed to be due to the voids being formed during the wiredrawing or rolling process used to produce the Al-Si alloy wire or ribbon. During the wiredrawing or rolling process, the hard Si phase is inhibited from deforming, while the Al phase surrounding the Si phase readily undergoes plastic deformation, allowing voids to form around the Si phase. Furthermore, regions with dispersed voids have the effect of alleviating stress and strain within the wire or ribbon during high-speed TCT temperature changes. By forming voids close to the Si phase, the Si phase not only effectively reduces thermal expansion, but also buffers thermal stress associated with temperature changes, further enhancing the effects of thermal expansion reduction and thermal stress reduction. As the upper limit temperature of the high-speed temperature cycle test increases, the thermal expansion difference between the wire or ribbon and the semiconductor increases, making the thermal stress buffering function of voids even more important at higher temperatures.

[0029] Furthermore, the relatively small size of the voids improves strain propagation locally, enhancing the effect of mitigating crack propagation within the electrode. If the voids grow large or propagate along cracks, there is concern that they may cause wire breakage during the wire drawing or rolling process. Therefore, with regard to voids related to joint reliability, we focus on voids with a sphere-equivalent diameter of 1 μm or more but less than 10 μm as determined by X-ray CT analysis.

[0030] A gradient structure with fewer voids in the surface of the wire or ribbon compared to the core is advantageous when temperature changes are severe, and is highly effective in improving the bonding life of high-speed TCT. This is thought to be because when there are fewer voids in the surface region, the effect of reducing the thermal expansion of the Si phase and the effect of reducing thermal strain are more efficiently transmitted to the electrode side of the bonding interface. Although the detailed mechanism is unknown, it is assumed that the voids absorb some of the thermal strain due to rapid temperature changes.

[0031] Specifically, the cross section is divided into three parts, distinguishing them from the center into a core part, an intermediate part, and a surface part, and the volume fraction of voids in each part is compared. That is, S0 is a closed curve representing the periphery of a cross section perpendicular to the central axis of the Al bonding wire or Al bonding ribbon, S1 is a closed curve that is 2 / 3 times similar to S0, and S2 is a closed curve that is 1 / 3 times similar to S0. S0, S1, and S2 are arranged so that their centers of gravity coincide with the center of gravity of the cross section perpendicular to the central axis of the Al bonding wire or Al bonding ribbon, and the region surrounded by S0 and S1 is the surface part, and the region surrounded by S2 is the core part. Here, "S1 is a closed curve that is 2 / 3 times similar to S0" means that S0 and S1 are similar figures with a similarity ratio of 3:2. Similarly, "S2 is a closed curve that is 1 / 3 times similar to S0" means that S0 and S2 are similar figures with a similarity ratio of 3:1. Figure 5 shows the relationship between S0, S1, and S2, as well as the core, intermediate, and surface regions, for Al bonding wire. In the case of Al bonding wire, if the radius of the cross section perpendicular to the central axis of the Al bonding wire is R, the circular region inside the concentric circle of radius 1 / 3R is the core, and the hollow circular region surrounded by the concentric circle of radius 2 / 3R and the concentric circle of radius R is the surface. Figure 6 also shows the relationship between S0, S1, and S2, as well as the core, intermediate, and surface regions, for Al bonding ribbon.

[0032] When the volume fraction of voids in the core portion having a sphere-equivalent diameter of 1 μm or more and less than 10 μm measured by X-ray CT analysis is Rd, and the volume fraction of voids in the surface portion having a sphere-equivalent diameter of 1 μm or more and less than 10 μm is Rf, the void gradient ratio (Rf / Rd) is 0.005 or more and 0.50 or less.

[0033] To calculate the void volume ratio, first calculate Vs, the volume of voids with a sphere-equivalent diameter of 1 μm or more but less than 10 μm, Vg, the total volume of voids of all sizes, and Vm, the volume of material. Next, calculate the ratio (Vs / (Vg+Vm)) by dividing Vs by the total volume of voids and material (Vg+Vm), to obtain the void volume ratio. Following this method, calculate Rd and Rf, the volume ratios of voids in the core and surface regions, respectively.

[0034] The following explains why a void gradient ratio (Rf / Rd) of 0.005 or more and 0.50 or less exhibits excellent high-speed temperature cycle reliability even in high-speed temperature cycle tests with high upper limit temperatures. A void gradient ratio (Rf / Rd) of 0.005 or more can prevent adverse effects caused by an excessively large difference in void volume fraction between the surface and core regions, and industrial manufacturing becomes relatively easy. Furthermore, a void gradient ratio (Rf / Rd) of 0.50 or less can sufficiently increase the difference in void volume fraction between the surface and core regions, resulting in sufficient improvement in high-speed temperature cycle tests with high upper limit temperatures.

[0035] The inventors have also confirmed that the Al bonding wire or Al bonding ribbon of the present invention exhibits an excellent shear strength multiplier. When evaluating the bondability of a first bonded portion, if the bonding wire or ribbon breaks internally during a shear test, it is difficult to accurately evaluate the bondability of high-strength materials. In this regard, a method for evaluating bond strength while suppressing the influence of material strength is to determine the shear strength as a value normalized by the material strength. After investigating the relationship between material strength and bondability, the inventors have found that calculating the ratio (SH / PS, hereinafter referred to as the "shear strength multiplier") obtained by dividing the shear strength (SH) by the 0.2% proof stress (PS) in a tensile test is effective. The factors governing the apparent shear strength of a first bonded portion can be divided into net bond strength and the deformation resistance of the wire or ribbon. The latter, deformation resistance, correlates with the yield strength at which the wire or ribbon yields and begins plastic deformation during a shear test. Therefore, in order to evaluate the net joint strength, we devised a method to exclude the influence of the yield strength of the wire or ribbon from the apparent shear strength. As the yield strength, we used the 0.2% proof stress, which is the strength value when the elongation rate in the tensile test is 0.2%.

[0036] That is, the inventors have found that the bond condition can be evaluated more accurately by using the shear strength ratio (SH / PS) normalized by the 0.2% proof stress (PS). A high shear strength ratio indicates that a good bond has been obtained at the bond interface. Using the shear strength ratio (SH / PS) as an index, a wire or ribbon with a high shear strength ratio can achieve both good bond strength and reduced bond damage, and also exhibit excellent performance in temperature cycle reliability.

[0037] From the viewpoint of exhibiting excellent high-speed temperature cycle reliability even in a high-speed temperature cycle test with a high upper limit temperature and from the viewpoint of exhibiting excellent shear strength magnification, the lower limit of the void gradient ratio (Rf / Rd) is 0.005 or more, preferably 0.008 or more, 0.010 or more, 0.015 or more, 0.020 or more, 0.030 or more, or 0.050 or more, more preferably 0.080 or more, or 0.10 or more, even more preferably 0.12 or more, or 0.15 or more, and particularly preferably 0.18 or more, or 0.20 or more. From the viewpoint of exhibiting excellent high-speed temperature cycle reliability even in a high-speed temperature cycle test with a high upper limit temperature and from the viewpoint of exhibiting excellent shear strength magnification, the upper limit of the void gradient ratio (Rf / Rd) is 0.50 or less, preferably 0.48 or less, or 0.45 or less, more preferably 0.42 or less, or 0.40 or less, even more preferably 0.38 or less, or 0.35 or less, and particularly preferably 0.32 or less, or 0.30 or less.

[0038] -Si concentration- A Si concentration in the range of 3.0% by mass to 20.0% by mass helps improve the shear strength ratio and reduce thermal distortion in the joint, thereby improving high-speed temperature cycle reliability. Specifically, a Si concentration of 3.0% by mass or more can significantly improve high-speed temperature cycle reliability. Furthermore, with advances and optimization of the equipment and conditions used in the manufacture and bonding of wires, etc., higher Si concentrations have become tolerable while suppressing defects such as wire breakage during processing, deterioration of surface properties, reduced initial joint strength due to hardening, and damage to semiconductor chips. However, a Si concentration of 20.0% by mass or less can effectively suppress these defects while improving the shear strength ratio. From the viewpoint of obtaining good high-speed temperature cycle reliability, the Si concentration in the Al bonding wire or Al bonding ribbon of the present invention is 3.0 mass% or more, preferably 3.5 mass% or more, more preferably 4.0 mass% or more, and even more preferably 4.2 mass% or more, 4.4 mass% or more, 4.5 mass% or more, 4.6 mass% or more, 4.8 mass% or more, or 5.0 mass% or more. In addition, from the viewpoint of improving the shear strength ratio while well suppressing defects such as a decrease in initial bond strength due to hardening and damage to the semiconductor chip, the Si concentration in the Al bonding wire or Al bonding ribbon of the present invention is 20.0 mass% or less, preferably 19.0 mass% or less, 18.0 mass% or less, 17.0 mass% or less, 16.0 mass% or less, 15.0 mass% or less, 14.5 mass% or less, 14.0 mass% or less, 13.5 mass% or less, 13.0 mass% or less, or 12.5 mass% or less. Furthermore, if the hardness of the Al bonding wire or Al bonding ribbon is high, damage to the semiconductor chip is more likely to occur during the first bonding depending on the bonding conditions of ultrasonic vibration and load. From the viewpoint of obtaining good bonding strength under a wider range of bonding conditions, the Si concentration in the Al bonding wire or Al bonding ribbon of the present invention is more preferably 12.0 mass% or less, even more preferably 11.5 mass% or less or 11.0 mass% or less, and particularly preferably 10.8 mass% or less, 10.6 mass% or less, 10.5 mass% or less, 10.4 mass% or less, 10.2 mass% or less, or 10.0 mass% or less.

[0039] For example, an ICP (Inductively Coupled Plasma) optical emission spectrometer or an ICP mass spectrometer can be used to analyze the concentration of elements contained in the Al bonding wire or Al bonding ribbon of the present invention. If elements derived from atmospheric contaminants such as oxygen or carbon are adsorbed on the surface of the Al bonding wire or Al bonding ribbon, it is effective to clean the surface with an acid or alkali depending on the adsorbed substance before analysis.

[0040] -Analysis of voids using X-ray CT- In the present invention, the void gradient ratio (Rf / Rd) is measured using X-ray CT analysis. X-ray CT analysis technology is excellent for non-destructively observing voids in materials, and by performing three-dimensional image analysis based on the data, it is possible to observe minute voids in metals with high precision. Below, an example of an analysis method for measuring voids in an Al bonding wire with a wire diameter of 300 μm is shown. The analysis conditions are not limited to these, and appropriate analysis conditions can be selected depending on the device and sample. For example, a 3D X-ray microscope "Xradia 520 Versa" (manufactured by ZEISS) can be used as an X-ray CT device. The main measurement conditions for X-ray CT are as follows: X-ray voltage 40 kV, X-ray output 3 W, wavelength limiting filter LE1, and magnification lens 4x. It has been confirmed that a transmission image with a transmittance of 30% to 70% can be obtained under the above conditions. To observe minute voids, the pixel size is set to approximately 0.7 μm, allowing the measurement field of view to cover an area of ​​approximately 700 μm x 700 μm x 700 μm. The sample length is set to approximately 10 mm, and the end of the sample is fixed. CT projection images can be measured by passing X-rays through the sample at a rotation angle of 360 degrees around the axis of the sample. After CT projection images are obtained at all angles, reconstruction processing such as center shifting can be performed to obtain 3D image data.

[0041] Next, the 3D image data is analyzed using image analysis software. Avizo Inspection can be used as the image analysis software. Specifically, a brightness histogram (Figure 3) is created using the measurement data. The brightness value at the midpoint of two peak positions corresponding to voids or external space and material is set as a threshold. Image data areas showing brightness lower than the threshold are identified as voids or external space (the environment surrounding the sample). Next, for the identified voids and external space, the external space areas are selected and removed from the image data area. This allows image data containing only voids in the material within the measurement area of ​​the sample to be identified. Next, using a similar process, image data with brightness higher than the threshold is identified as image data related to material. This "material" refers to something that has mass and volume. In the case of the Al bonding wire or Al bonding ribbon of the present invention, this includes Al, Si, element group 1, element group 2, and other elements, as well as their alloys, oxides, and intermetallic compounds. Furthermore, in the X-ray CT analysis of the present invention, pixels showing brightness higher than the above threshold are considered to contain material.

[0042] The image data of the voids and material thus obtained are analyzed. The method for calculating the volume fraction of voids with a sphere-equivalent diameter of 1 μm or more and less than 10 μm is now specifically explained. Image analysis software is used to calculate the total volume Vs of voids with a sphere-equivalent diameter of 1 μm or more and less than 10 μm, the total volume Vg of voids of all sizes, and the volume Vm of the material. The ratio (Vs / (Vg+Vm)) obtained by dividing Vs by the total volume of voids and material (Vg+Vm) is the void volume fraction. Using this method, the volume fractions Rd and Rf of voids in the core and surface regions are calculated.

[0043] -Distribution of circle equivalent diameter of Si phase in L cross section- In order to obtain excellent high-speed temperature cycle reliability even when the number of cycles is increased in a high-speed TCT with a high upper limit temperature, when the number of Si phases having a circle-equivalent diameter of 0.5 μm or more and 0.8 μm or less in the L cross section of the Al bonding wire or Al bonding ribbon is Ns, and the number of Si phases having a circle-equivalent diameter of 0.5 μm or more in the L cross section is Nc, the ratio of Ns to Nc [Ns / Nc × 100(%)] (hereinafter also referred to as the "small diameter ratio of Si phases") is 30% or more and 95% or less. The lower limit of the small diameter ratio of Si phases [Ns / Nc × 100(%)] is preferably 32% or more, 35% or more, 38% or more, 40% or more, 42% or more, or 45% or more, more preferably 48% or more, even more preferably 50% or more or 52% or more, and particularly preferably 55% or more or 60% or more. The upper limit of the small diameter ratio of the Si phase [Ns / Nc × 100(%)] is preferably 92% or less, 90% or less, 88% or less, or 85% or less, more preferably 82% or less, or 80% or less, even more preferably 78% or less, or 75% or less, and particularly preferably 72% or less, or 70% or less. In one embodiment, the small diameter ratio of the Si phase [Ns / Nc × 100(%)] is preferably 40% or more and 90% or less.

[0044] The reason why the proportion of Si phases with equivalent circle diameters of 0.5 μm to 0.8 μm is important is believed to be as follows: Specifically, Si phases with equivalent circle diameters of 0.5 μm or more have a sufficiently large volume, which allows them to sufficiently reduce thermal expansion. Furthermore, from the perspective of the analytical accuracy of current EDS and EBSD analysis equipment, it is appropriate to target Si phases with equivalent circle diameters of 0.5 μm or more. On the other hand, Si phases with equivalent circle diameters of 0.8 μm or less sufficiently uniformize stress and strain near the bonding interface due to the Si phase. Furthermore, the reason why a small diameter ratio of the Si phase [Ns / Nc × 100 (%)] in the range of 30% to 95% can achieve excellent high-speed temperature cycle reliability even with a longer cycle number in high-speed TCT with a high upper limit temperature is believed to be as follows: Specifically, when the small diameter ratio of the Si phase is 30% or more, there is a sufficient amount of Si phases with small equivalent circle diameters, which uniformly distributes thermal strain and stabilizes high-speed temperature cycle reliability. On the other hand, when the small diameter ratio of the Si phase is 95% or less, the total volume of the Si phase increases, and the effect of reducing thermal expansion of the entire bonding region can be maintained at a high level, thereby sufficiently increasing the effect of improving high-speed temperature cycle reliability. In one embodiment, when the void gradient ratio (Rf / Rd) of the Al bonding wire or Al bonding ribbon of the present invention is 0.005 or more and 0.50 or less, the Al bonding wire or Al bonding ribbon exhibits excellent high-speed temperature cycle reliability in a high-speed temperature cycle test with 10,000 cycles, while when the small diameter ratio of the Si phase of the Al bonding wire or Al bonding ribbon is 30% or more and 95% or less, the Al bonding wire or Al bonding ribbon also exhibits excellent high-speed temperature cycle reliability in a high-speed temperature cycle test with 13,000 cycles.

[0045] Here, each Si phase is formed in a particle shape, and it has been confirmed that the number of Si phase particles has a large effect on thermal distortion at the bonding interface. Therefore, the ratio of the number of Si phase particles can be used to determine high-speed temperature cycling reliability. However, because the effect of coarse particles is overestimated by particle area, it is difficult to accurately evaluate the correlation with high-speed temperature cycling reliability.

[0046] -Method for measuring the circle equivalent diameter of the Si phase and calculating the small diameter ratio- This paper describes a method for measuring the circle-equivalent diameter of the Si phase in the L-section of an Al bonding wire or Al bonding ribbon. The circle-equivalent diameter of the Si phase in the L-section can be measured using a SEM-EDS-EBSD instrument. Specifically, this method combines information on the Al and Si concentrations obtained by SEM-EDS (Scanning Electron Microscope-Energy Dispersive X-ray Spectroscopy) with information on the crystal orientation obtained by electron backscatter diffraction (EBSD). More specifically, in the measurement area where the L-section of the Al bonding wire or Al bonding ribbon is used as the inspection surface, Al and Si concentration measurements are performed using EDS, and crystal orientation analysis is performed using EBSD simultaneously. Next, the Al and Si phases are separated and extracted from the EDS measurement results using the analysis software provided with the instrument. Specifically, it is preferable to use the Chi Scan function, which is a function of the analysis software OIM Data Collection or OIM Analysis (both manufactured by TSL Solutions) attached to an FE-SEM (Field Emission-Scanning Electron Microscope) device. Then, for the region identified as the Si phase, the crystal orientation can be analyzed using the analysis software attached to the device. If the orientation difference between the measurement points is 15° or more, it is determined to be a grain boundary and the circle equivalent diameter is calculated. In the process of calculating the small diameter ratio of the Si phase, parts where the crystal orientation cannot be measured or parts where the crystal orientation can be measured but the reliability of the orientation analysis is low are excluded from the calculation. Therefore, in one embodiment, the small diameter ratio of the Si phase in the L cross section of the Al bonding wire or Al bonding ribbon of the present invention is calculated by the following steps (1) to (3). (1) The L-section of the Al bonding wire or Al bonding ribbon is used as the inspection surface, and the Al and Si concentrations are measured using EDS and the crystal orientation is measured using EBSD simultaneously. (2) Use the Chi Scan function to separate and extract Al and Si. Specifically, by setting a tolerance equivalent to the Si threshold from the Si EDS measurement results, Al and Si can be separated and identified. The crystal orientation can be analyzed using the Al and Si crystal information in the material file. (3) For the areas identified as Si phase, the crystal orientation is analyzed, and if the misorientation between measurement points is 15° or more, it is determined to be a grain boundary, and the circle-equivalent diameter of each crystal grain is calculated. The number of crystal grains identified as Si phase is tallied to determine the total number of Si phase particles, Nc. Here, Si phases with a circle-equivalent diameter of 0.5 μm or more are targeted. Taking into account the analytical accuracy of current ESD and EBSD analysis equipment, fine particles less than 0.5 μm are excluded. In addition, the number of Si phases, Ns, with a circle-equivalent diameter in the range of 0.5 μm to 0.8 μm is tallied. Then, the ratio of Ns to Nc [Ns / Nc × 100 (%)] (the small diameter ratio of Si phases) is calculated.

[0047] In step (2) above, the Tolerance (%) setting can be selected in the range of 20 to 40%, and for standard analysis of the L cross section of Al bonding wire or Al bonding ribbon, it is preferable to compare at approximately 30%. Here is a supplementary explanation of the procedure for adjusting this Tolerance. It is preferable to select or confirm the Tolerance value so that the shape and size of the Si phase extracted and identified by the Chi Scan function are equivalent to those identified in the EDS map, which displays the Si element concentration in EDS analysis in two dimensions.

[0048] In the present invention, the small diameter ratio of the Si phase is defined as the average (arithmetic mean) of the orientation ratio values ​​obtained by measuring at three or more locations. When selecting the measurement area, in order to ensure the objectivity of the measurement data, it is preferable to obtain measurement samples from the Al bonding wire or Al bonding ribbon to be measured at intervals of 50 cm or more along the central axis of the Al bonding wire or Al bonding ribbon and provide them for measurement. Furthermore, in the present invention, the measurement area for crystal orientation using the EBSD method has a length in the direction of the central axis of the Al bonding wire or Al bonding ribbon of 300 μm or more but less than 800 μm, and it is desirable that the entire Al bonding wire or Al bonding ribbon is included in the direction perpendicular to the central axis of the Al bonding wire or Al bonding ribbon. However, if the size is too large to measure the entire area, it can be adjusted to a range of less than 600 μm.

[0049] An example of the measurement results is shown in Figure 4. The horizontal axis shows the circle equivalent diameter of the Si phase, with each interval having a width of 0.1 μm, and the vertical axis shows the number of particles. The range of 0.5 μm to 0.8 μm is shown with a double-headed arrow, and the particle number ratio [Ns / Nc × 100 (%)] in this interval is 50%. In a high-speed temperature cycle test of this Al bonding wire, the decrease in strength was kept low even when the maximum temperature was raised to 185°C, confirming that it has good high-speed temperature cycle reliability.

[0050] -Addition of Sr, Na, Eu, and P- The Al bonding wire or Al bonding ribbon of the present invention may further contain one or more of Sr, Na, Eu, and P (hereinafter also referred to as the "first element group"). The total concentration of the first element group may be 0 ppm by mass, preferably 1 ppm by mass or more, more preferably 3 ppm by mass or more, even more preferably 5 ppm by mass or more, and particularly preferably 8 ppm by mass or more or 10 ppm by mass or more. The upper limit of the total concentration of the first element group is preferably 10,000 ppm by mass or less or 8,000 ppm by mass or less, more preferably 5,000 ppm by mass or less or 3,000 ppm by mass or less, even more preferably 2,000 ppm by mass or less or 1,000 ppm by mass or less, and particularly preferably 900 ppm by mass or less or 800 ppm by mass or less. In one embodiment, the total concentration of the first element group is preferably 10 ppm by mass or more and 800 ppm by mass or less.

[0051] The Al bonding wire or Al bonding ribbon of the present invention further contains one or more of Sr, Na, Eu, and P in a total amount of 10 mass ppm to 800 mass ppm, thereby reducing the frequency of wire breakage during wiredrawing of the Al bonding wire or Al bonding ribbon. Al alloys containing a high concentration of Si at 3.0 mass% to 20.0 mass% tend to have a higher frequency of wire breakage during the wiredrawing process. This is thought to be due in part to the fact that Si phase particles crystallized during solidification cause stress concentration during wiredrawing, inducing wire breakage. It is presumed that the addition of the first element group can uniformly distribute the particulate Si phase and inhibit the growth and coarsening of the Si phase, thereby alleviating stress concentration during wiredrawing and reducing wire breakage. It is thought that the addition of the first element group, along with controlling the void gradient ratio, enhances the effect of alleviating stress concentration during wiredrawing.

[0052] From the viewpoint of reducing the frequency of wire breakage during wire drawing, the total concentration of the first element group in the Al bonding wire or Al bonding ribbon of the present invention is more preferably 20 ppm by mass or more, even more preferably 30 ppm by mass or more, 40 ppm by mass or more, or 50 ppm by mass or more, and the upper limit is preferably 750 ppm by mass or less, more preferably 740 ppm by mass or less, 720 ppm by mass or less, or 700 ppm by mass or less, even more preferably 680 ppm by mass or less, 650 ppm by mass or less, 620 ppm by mass or less, or 600 ppm by mass or less, and particularly preferably 580 ppm by mass or less, 550 ppm by mass or less, 520 ppm by mass or less, or 500 ppm by mass or less.

[0053] When the Al bonding wire or Al bonding ribbon of the present invention contains one or more elements from the first element group, it may contain one element from the first element group, two elements from the first element group, three elements from the first element group, or all four elements from the first element group. Also, when the Al bonding wire or Al bonding ribbon of the present invention contains one or more elements from the first element group, it may contain Sr, Na, Eu, or P.

[0054] When the Al bonding wire or Al bonding ribbon of the present invention contains Sr from the first element group, the Sr concentration may be 0 ppm by mass, preferably 1 ppm by mass or more, 3 ppm by mass or more, 5 ppm by mass or more, or 8 ppm by mass or more. Furthermore, from the viewpoint of reducing the frequency of wire breakage during wire drawing, the Sr concentration is more preferably 10 ppm by mass or more, and even more preferably 20 ppm by mass or more, 30 ppm by mass or more, 40 ppm by mass or more, or 50 ppm by mass or more. The upper limit of the Sr concentration is preferably 10,000 ppm by mass or less, 8,000 ppm by mass or less, 5,000 ppm by mass or less, 3,000 ppm by mass or less, 2,000 ppm by mass or less, 1,000 ppm by mass or less, or 900 ppm by mass or less. Furthermore, from the viewpoint of reducing the frequency of wire breakage during wiredrawing, the Sr concentration is more preferably 800 ppm by mass or less, even more preferably 750 ppm by mass or less, 740 ppm by mass or less, 720 ppm by mass or less, 700 ppm by mass or less, 680 ppm by mass or less, 650 ppm by mass or less, 620 ppm by mass or less, 600 ppm by mass or less, 580 ppm by mass or less, 550 ppm by mass or less, 520 ppm by mass or less, or 500 ppm by mass or less.

[0055] When the Al bonding wire or Al bonding ribbon of the present invention contains Na in the first element group, the Na concentration may be 0 ppm by mass, preferably 1 ppm by mass or more, 3 ppm by mass or more, 5 ppm by mass or more, or 8 ppm by mass or more. Furthermore, from the viewpoint of reducing the frequency of wire breakage during wire drawing, the Na concentration is more preferably 10 ppm by mass or more, even more preferably 20 ppm by mass or more, 30 ppm by mass or more, 40 ppm by mass or more, or 50 ppm by mass or more. The upper limit of the Na concentration is preferably 10,000 ppm by mass or less, 8,000 ppm by mass or less, 5,000 ppm by mass or less, 3,000 ppm by mass or less, 2,000 ppm by mass or less, 1,000 ppm by mass or less, or 900 ppm by mass or less. Furthermore, from the viewpoint of reducing the frequency of wire breakage during wiredrawing, the Na concentration is more preferably 800 ppm by mass or less, even more preferably 750 ppm by mass or less, 740 ppm by mass or less, 720 ppm by mass or less, 700 ppm by mass or less, 680 ppm by mass or less, 650 ppm by mass or less, 620 ppm by mass or less, 600 ppm by mass or less, 580 ppm by mass or less, 550 ppm by mass or less, 520 ppm by mass or less, or 500 ppm by mass or less.

[0056] When the Al bonding wire or Al bonding ribbon of the present invention contains Eu from the first element group, the Eu concentration may be 0 ppm by mass, preferably 1 ppm by mass or more, 3 ppm by mass or more, 5 ppm by mass or more, or 8 ppm by mass or more. Furthermore, from the viewpoint of reducing the frequency of wire breakage during wire drawing, the Eu concentration is more preferably 10 ppm by mass or more, even more preferably 20 ppm by mass or more, 30 ppm by mass or more, 40 ppm by mass or more, or 50 ppm by mass or more. The upper limit of the Eu concentration is preferably 10,000 ppm by mass or less, 8,000 ppm by mass or less, 5,000 ppm by mass or less, 3,000 ppm by mass or less, 2,000 ppm by mass or less, 1,000 ppm by mass or less, or 900 ppm by mass or less. Furthermore, from the viewpoint of reducing the frequency of wire breakage during wiredrawing, the Eu concentration is more preferably 800 ppm by mass or less, even more preferably 750 ppm by mass or less, 740 ppm by mass or less, 720 ppm by mass or less, 700 ppm by mass or less, 680 ppm by mass or less, 650 ppm by mass or less, 620 ppm by mass or less, 600 ppm by mass or less, 580 ppm by mass or less, 550 ppm by mass or less, 520 ppm by mass or less, or 500 ppm by mass or less.

[0057] When the Al bonding wire or Al bonding ribbon of the present invention contains P from the first element group, the P concentration may be 0 ppm by mass, preferably 1 ppm by mass or more, 3 ppm by mass or more, 5 ppm by mass or more, or 8 ppm by mass or more. Furthermore, from the viewpoint of reducing the frequency of wire breakage during wire drawing, the P concentration is more preferably 10 ppm by mass or more, even more preferably 20 ppm by mass or more, 30 ppm by mass or more, 40 ppm by mass or more, or 50 ppm by mass or more. The upper limit of the P concentration is preferably 10,000 ppm by mass or less, 8,000 ppm by mass or less, 5,000 ppm by mass or less, 3,000 ppm by mass or less, 2,000 ppm by mass or less, 1,000 ppm by mass or less, or 900 ppm by mass or less. Furthermore, from the viewpoint of reducing the frequency of wire breakage during wiredrawing, the P concentration is more preferably 800 ppm by mass or less, even more preferably 750 ppm by mass or less, 740 ppm by mass or less, 720 ppm by mass or less, 700 ppm by mass or less, 680 ppm by mass or less, 650 ppm by mass or less, 620 ppm by mass or less, 600 ppm by mass or less, 580 ppm by mass or less, 550 ppm by mass or less, 520 ppm by mass or less, or 500 ppm by mass or less.

[0058] -Addition of Ni, Ti, Fe, and Mg- The Al bonding wire or Al bonding ribbon of the present invention may further contain one or more of Ni, Ti, Fe, and Mg (hereinafter also referred to as the "second element group"). The total concentration of the second element group may be 0 ppm by mass, preferably 1 ppm by mass or more or 3 ppm by mass or more, more preferably 5 ppm by mass or more or 8 ppm by mass or more, even more preferably 10 ppm by mass or more or 30 ppm by mass or more, particularly preferably 50 ppm by mass or more, 80 ppm by mass or more, or 100 ppm by mass or more. The upper limit of the total concentration of the second element group is preferably 10,000 ppm by mass or less, more preferably 8,000 ppm by mass or less, even more preferably 5,000 ppm by mass or less, particularly preferably 3,000 ppm by mass or less, 2,000 ppm by mass or less, 1,800 ppm by mass or less, 1,600 ppm by mass or less, or 1,500 ppm by mass or less. In one embodiment, the total concentration of the second element group is preferably 100 ppm by mass or more and 1,500 ppm by mass or less.

[0059] By further containing at least one of Ni, Ti, Fe, and Mg in a total amount of 100 ppm by mass to 1500 ppm by mass, the Al bonding wire or Al bonding ribbon of the present invention can suppress scratches and abrasions on the surface of the Al bonding wire or Al bonding ribbon, resulting in a smooth surface. Al alloys containing Si at a high concentration of 3.0% by mass to 20.0% by mass can harden the surface and cause the Si phase and Al oxide present on the surface to fall off, resulting in scratches and abrasions on the surface during wiredrawing, resulting in an Al bonding wire or Al bonding ribbon with significant surface irregularities. It is believed that the addition of the second element group stabilizes the Al oxide on the surface of the Al bonding wire or Al bonding ribbon, refines the structure of the Al crystal grains, and hardens them, thereby reducing scratches and abrasions during wiredrawing. It is believed that by controlling the void gradient rate and adding the second element group, the occurrence of scratches and abrasions on the surface of the Al bonding wire or Al bonding ribbon can be suppressed, thereby enhancing the effect of forming a smooth surface.

[0060] From the viewpoint of forming an Al bonding wire or Al bonding ribbon having a smooth surface by suppressing the occurrence of scratches and abrasions on the surface, the total concentration of the second element group in the Al bonding wire or Al bonding ribbon of the present invention is more preferably 150 mass ppm or more, even more preferably 200 mass ppm or more, 250 mass ppm or more, or 300 mass ppm or more, and the upper limit is more preferably 1200 mass ppm or less, even more preferably 1000 mass ppm or less, 900 mass ppm or less, or 800 mass ppm or less, and particularly preferably 700 mass ppm or less, 600 mass ppm or less, or 500 mass ppm or less.

[0061] When the Al bonding wire or Al bonding ribbon of the present invention contains one or more elements from the second element group, it may contain one element from the second element group, two elements from the second element group, three elements from the second element group, or all four elements from the second element group. Also, when the Al bonding wire or Al bonding ribbon of the present invention contains one or more elements from the second element group, it may contain Ni, Ti, Fe, or Mg.

[0062] When the Al bonding wire or Al bonding ribbon of the present invention contains Ni from the second element group, the Ni concentration may be 0 ppm by mass, preferably 1 ppm by mass or more, 3 ppm by mass or more, 5 ppm by mass or more, 8 ppm by mass or more, 10 ppm by mass or more, 30 ppm by mass or more, 50 ppm by mass or more, or 80 ppm by mass or more. Furthermore, from the viewpoint of suppressing the occurrence of scratches and scraping on the surface and forming an Al bonding wire or Al bonding ribbon with a smooth surface, the Ni concentration is more preferably 100 ppm by mass or more, even more preferably 150 ppm by mass or more, 200 ppm by mass or more, 250 ppm by mass or more, or 300 ppm by mass. The upper limit of the Ni concentration is preferably 10,000 ppm by mass or less, 8,000 ppm by mass or less, 5,000 ppm by mass or less, or 3,000 ppm by mass or less. Furthermore, from the viewpoint of suppressing the occurrence of scratches and abrasions on the surface and forming an Al bonding wire or Al bonding ribbon with a smooth surface, the Ni concentration is more preferably 2000 mass ppm or less, even more preferably 1800 mass ppm or less, 1600 mass ppm or less, 1500 mass ppm or less, 1200 mass ppm or less, 1000 mass ppm or less, 900 mass ppm or less, 800 mass ppm or less, 700 mass ppm or less, 600 mass ppm or less, or 500 mass ppm or less.

[0063] When the Al bonding wire or Al bonding ribbon of the present invention contains Ti from the second element group, the Ti concentration may be 0 ppm by mass, preferably 1 ppm by mass or more, 3 ppm by mass or more, 5 ppm by mass or more, 8 ppm by mass or more, 10 ppm by mass or more, 30 ppm by mass or more, 50 ppm by mass or more, or 80 ppm by mass or more. Furthermore, from the viewpoint of suppressing the occurrence of scratches and scraping on the surface and forming an Al bonding wire or Al bonding ribbon with a smooth surface, the Ti concentration is more preferably 100 ppm by mass or more, even more preferably 150 ppm by mass or more, 200 ppm by mass or more, 250 ppm by mass or more, or 300 ppm by mass. The upper limit of the Ti concentration is preferably 10,000 ppm by mass or less, 8,000 ppm by mass or less, 5,000 ppm by mass or less, or 3,000 ppm by mass or less. Furthermore, from the viewpoint of suppressing the occurrence of scratches and abrasions on the surface and forming an Al bonding wire or Al bonding ribbon with a smooth surface, the Ti concentration is more preferably 2000 mass ppm or less, even more preferably 1800 mass ppm or less, 1600 mass ppm or less, 1500 mass ppm or less, 1200 mass ppm or less, 1000 mass ppm or less, 900 mass ppm or less, 800 mass ppm or less, 700 mass ppm or less, 600 mass ppm or less, or 500 mass ppm or less.

[0064] When the Al bonding wire or Al bonding ribbon of the present invention contains Fe from the second element group, the Fe concentration may be 0 ppm by mass, preferably 1 ppm by mass or more, 3 ppm by mass or more, 5 ppm by mass or more, 8 ppm by mass or more, 10 ppm by mass or more, 30 ppm by mass or more, 50 ppm by mass or more, or 80 ppm by mass or more. Furthermore, from the viewpoint of suppressing the occurrence of scratches and scraping on the surface and forming an Al bonding wire or Al bonding ribbon with a smooth surface, the Fe concentration is more preferably 100 ppm by mass or more, even more preferably 150 ppm by mass or more, 200 ppm by mass or more, 250 ppm by mass or more, or 300 ppm by mass. The upper limit of the Fe concentration is preferably 10,000 ppm by mass or less, 8,000 ppm by mass or less, 5,000 ppm by mass or less, or 3,000 ppm by mass or less. Furthermore, from the viewpoint of suppressing the occurrence of scratches and abrasions on the surface and forming an Al bonding wire or Al bonding ribbon with a smooth surface, the Fe concentration is more preferably 2000 mass ppm or less, even more preferably 1800 mass ppm or less, 1600 mass ppm or less, 1500 mass ppm or less, 1200 mass ppm or less, 1000 mass ppm or less, 900 mass ppm or less, 800 mass ppm or less, 700 mass ppm or less, 600 mass ppm or less, or 500 mass ppm or less.

[0065] When the Al bonding wire or Al bonding ribbon of the present invention contains Mg from the second element group, the Mg concentration may be 0 ppm by mass, preferably 1 ppm by mass or more, 3 ppm by mass or more, 5 ppm by mass or more, 8 ppm by mass or more, 10 ppm by mass or more, 30 ppm by mass or more, 50 ppm by mass or more, or 80 ppm by mass or more. Furthermore, from the viewpoint of suppressing the occurrence of scratches and scraping on the surface and forming an Al bonding wire or Al bonding ribbon with a smooth surface, the Mg concentration is more preferably 100 ppm by mass or more, even more preferably 150 ppm by mass or more, 200 ppm by mass or more, 250 ppm by mass or more, or 300 ppm by mass. The upper limit of the Mg concentration is preferably 10,000 ppm by mass or less, 8,000 ppm by mass or less, 5,000 ppm by mass or less, or 3,000 ppm by mass or less. Furthermore, from the viewpoint of suppressing the occurrence of scratches and abrasions on the surface and forming an Al bonding wire or Al bonding ribbon with a smooth surface, the Mg concentration is more preferably 2000 mass ppm or less, even more preferably 1800 mass ppm or less, 1600 mass ppm or less, 1500 mass ppm or less, 1200 mass ppm or less, 1000 mass ppm or less, 900 mass ppm or less, 800 mass ppm or less, 700 mass ppm or less, 600 mass ppm or less, or 500 mass ppm or less.

[0066] As the aluminum raw material for manufacturing the Al bonding wire or Al bonding ribbon of the present invention, it is preferable to use Al with a purity of 4N (Al: 99.99% by mass or more), and it is even more preferable to use Al with a lower impurity content of 5N (Al: 99.999% by mass or more). In one embodiment, Al with a purity of 3N (Al: 99.9% by mass or more) may be used.

[0067] The Al bonding wire or Al bonding ribbon of the present invention may further contain elements other than Al, Si, the first element group, and the second element group (hereinafter also referred to as "other elements"). That is, "other elements" are elements other than Al, Si, Sr, Na, Eu, P, Ni, Ti, Fe, and Mg, and the Al bonding wire or Al bonding ribbon of the present invention may further contain elements other than Al, Si, Sr, Na, Eu, P, Ni, Ti, Fe, and Mg. The total concentration of other elements in the Al bonding wire or Al bonding ribbon is not particularly limited as long as it does not impair the effects of the present invention. The total concentration of the other elements may be, for example, 0.5% by mass or less, 0.4% by mass or less, 0.3% by mass or less, 0.2% by mass or less, 0.15% by mass or less, 0.1% by mass or less, 0.08% by mass or less, 0.06% by mass or less, 0.05% by mass or less, 0.04% by mass or less, 0.03% by mass or less, 0.025% by mass or less, 0.02% by mass or less, 0.018% by mass or less, 0.016% by mass or less, 0.015% by mass or less, 0.014% by mass or less, 0.012% by mass or less, or 0.01% by mass or less. The lower limit of the total concentration of the other elements is not particularly limited, and may be 0% by mass.

[0068] In one embodiment, the remainder of the Al bonding wire or Al bonding ribbon of the present invention consists of Al and other elements. Therefore, in a preferred embodiment, the Al bonding wire or Al bonding ribbon of the present invention consists of Al, Si, and other elements. In another preferred embodiment, the Al bonding wire or Al bonding ribbon of the present invention consists of Al, Si, one or more elements from the first element group, and other elements. In yet another preferred embodiment, the Al bonding wire or Al bonding ribbon of the present invention consists of Al, Si, one or more elements from the second element group, and other elements. In yet another preferred embodiment, the Al bonding wire or Al bonding ribbon of the present invention consists of Al, Si, one or more elements from the first element group, one or more elements from the second element group, and other elements.

[0069] In one embodiment, the remainder of the Al bonding wire or Al bonding ribbon of the present invention consists of Al and unavoidable impurities. Thus, in a preferred embodiment, the Al bonding wire or Al bonding ribbon of the present invention consists of Al, Si, and unavoidable impurities. In another preferred embodiment, the Al bonding wire or Al bonding ribbon of the present invention consists of Al, Si, one or more elements from the first element group, and unavoidable impurities. In yet another preferred embodiment, the Al bonding wire or Al bonding ribbon of the present invention consists of Al, Si, one or more elements from the second element group, and unavoidable impurities. In yet another preferred embodiment, the Al bonding wire or Al bonding ribbon of the present invention consists of Al, Si, one or more elements from the first element group, one or more elements from the second element group, and unavoidable impurities.

[0070] In a preferred embodiment, the Al bonding wire or Al bonding ribbon of the present invention does not have a coating mainly composed of a metal other than Al on the outer periphery of the Al bonding wire or Al bonding ribbon. Here, "a coating mainly composed of a metal other than Al" refers to a coating in which the content of a metal other than Al is 50 mass % or more.

[0071] The Al bonding wire or Al bonding ribbon of the present invention may be either an Al bonding wire or an Al bonding ribbon. When the present invention is an Al bonding wire, its wire diameter is not particularly limited and may be, for example, 50 μm or more, 60 μm or more, 80 μm or more, 100 μm or more, 120 μm or more, 140 μm or more, 150 μm or more, 180 μm or more, or 200 μm or more. The upper limit of the wire diameter is not particularly limited and may be, for example, 600 μm or less, 550 μm or less, 500 μm or less, 450 μm or less, or 400 μm or less. In one embodiment, the wire diameter of the Al bonding wire of the present invention may be in the range of 100 to 600 μm, and preferably 200 to 400 μm. When the present invention is an Al bonding ribbon, the dimensions (width W x thickness T) of its rectangular or approximately rectangular cross section are not particularly limited, and for example, W may be 100 to 3000 μm, and T may be 50 to 600 μm.

[0072] The Al bonding wire or Al bonding ribbon of the present invention can provide excellent high-speed temperature cycle reliability even in high-speed temperature cycle tests with a high upper limit temperature. Therefore, the Al bonding wire or Al bonding ribbon of the present invention can be suitably used as an Al bonding wire or Al bonding ribbon for semiconductor devices. The Al bonding wire or Al bonding ribbon of the present invention can be particularly suitably used as an Al bonding wire or Al bonding ribbon for power semiconductor devices, and more suitably used as an Al bonding wire or Al bonding ribbon for SiC power semiconductor devices.

[0073] -Method for manufacturing Al bonding wire or Al bonding ribbon- An example of a method for manufacturing an Al bonding wire or Al bonding ribbon according to the present invention will be described below. Hereinafter, an example will be described in relation to the manufacture of an Al bonding wire.

[0074] The Al and alloying elements used as raw materials preferably have a high purity. Al preferably has a purity of 99.5% by mass or more, with the remainder consisting of inevitable impurities, more preferably a purity of 99.9% by mass or more, with the remainder consisting of inevitable impurities, and even more preferably a purity of 99.99% by mass or more, with the remainder consisting of inevitable impurities. The Si, first element group, second element group, and other elements used as alloying elements preferably have a purity of 99.9% by mass or more, with the remainder consisting of inevitable impurities, and more preferably a purity of 99.99% by mass or more, with the remainder consisting of inevitable impurities. Continuous casting can be used as a casting process for melting and solidifying the Al alloy used for Al bonding wire. In continuous casting, a molten liquid containing the Al raw material and the alloying element raw materials is poured into a water-cooled mold, while the cast material (ingot) is continuously drawn from below the mold. The atmosphere in the furnace during melting is preferably an inert or reducing atmosphere to prevent excessive oxidation of Al, Si, the first element group, the second element group, and other elements that make up the wire. The maximum temperature that the molten metal reaches during melting is preferably in the range of 800°C or higher and lower than 1050°C, taking into consideration factors such as ensuring the fluidity of the molten metal and making it easier to control the shape and size of the Si phase during solidification. Cooling methods after melting can include water cooling, furnace cooling, and air cooling.

[0075] The cylindrical ingot obtained by melting is subjected to solution treatment by heating at high temperature, and then repeatedly drawn using a die to produce wire of the desired diameter. After the drawing process, the wire is subjected to final heat treatment in an electric furnace and can be used as Al bonding wire.

[0076] To control the porosity gradient rate and the small diameter ratio of the Si phase, it is effective to control the heat treatment conditions such as solution treatment, homogenization treatment, and final heat treatment, as well as the wire drawing conditions. During wire drawing, it is effective to use a lubricant to ensure lubrication at the contact interface between the wire and the die.

[0077] <Control of void gradient ratio> In particular, in order to control the void gradient rate of the wire, it is effective to control the wire feed speed (wire drawing speed) in the wire drawing process, the die area reduction rate, lubricity, and the atmosphere in the intermediate annealing, as will be explained individually below.

[0078] Controlling the wire feed speed to a high speed according to the wire diameter used for wire drawing to assist deformation of the surface region is effective in promoting the void concentration gradient. As a specific example, in the process of drawing wire in a range from half the wire diameter at the start of wire drawing to the final wire diameter, it is preferable to set the average wire drawing speed to be 20 m / min or more and less than 50 m / min.

[0079] Regarding the die area reduction rate during wire drawing, it is effective to draw large diameter wires at a high area reduction rate and small diameter wires at a low area reduction rate. The high area reduction rate in the first half of the wire drawing concentrates the processing strain at the center of the wire, promoting an increase in voids in the core, while the low area reduction rate in the second half of the wire drawing gently deforms the area near the surface, making it possible to reduce voids in the surface area. As a specific example, it is preferable that the die area reduction rate from the wire diameter at the start of wire drawing to half of that diameter be in the range of 20% to less than 40%, and that from half the wire diameter to the final wire diameter be in the range of 10% to less than 25%. Here, if the wire area reduction rate per die is P1, P1 can be expressed by the following formula.

[0080] P1={(R2 2 -R1 2 ) / R2 2}×100 In the formula, R2 represents the diameter (mm) of the wire before processing, and R1 represents the diameter (mm) of the wire after processing.

[0081] In addition, the lubricity of the aqueous solution used in the wiredrawing process is improved during heat treatment at the final wire diameter or intermediate heat treatment at a small diameter close to the final wire diameter, promoting deformation that elongates the surface in the wiredrawing direction and thereby assisting in the concentration gradient of voids in the surface region.The lubricating liquid used is preferably an aqueous liquid containing a surfactant or the like that reduces the friction coefficient.

[0082] Performing intermediate heat treatment during wire drawing in an atmosphere of an inert gas such as N2 gas is also effective for controlling the concentration gradient of voids. It can assist in controlling the oxidation of Si in Al during wire drawing to keep the Si concentration near the surface low, and thus can reduce the void volume fraction near the surface.

[0083] <Control of the small-diameter ratio of the Si phase> Optimizing the intermediate heat treatment conditions and the final heat treatment conditions as a set is effective for controlling the small-diameter ratio of the Si phase. Specifically, when the temperature of the intermediate heat treatment at the wire diameter (final) closest to the final wire diameter is Tm (°C) and the temperature of the final heat treatment at the final wire diameter is Tc (°C), by making the temperature Tm of the intermediate heat treatment 50 °C or more higher than the temperature Tc of the final heat treatment, it becomes easy to adjust the small-diameter ratio of the Si phase to 30% or more and 95% or less. Specifically, by increasing the temperature of the intermediate heat treatment, the Si dissolved in the Al phase can be homogenized, and by lowering the temperature of the final heat treatment process, the solid solution concentration of Si in the Al phase can be decreased. By combining these temperatures, the number of minute Si phases can be increased, and the small-diameter ratio of the Si phase can be increased. Adjusting the temperature of the last intermediate heat treatment among multiple intermediate heat treatments is more effective, and it is considered that it can promote increasing the number of minute Si phases by utilizing the dislocations increased in the subsequent processing steps.

[0084] Regarding the final heat treatment conditions, it is effective to adjust within the temperature range of 200 °C or more and less than 360 °C and the time range of 2 hours or more and less than 20 hours. When recovery and recrystallization of the Al phase proceed due to the final heat treatment, at the same time, the amount of Si dissolved in the Al phase changes depending on the heat treatment temperature, and the recrystallization temperature changes. By adjusting the progress of recrystallization due to the final heat treatment, it becomes easy to control the small-diameter ratio of the Si phase. For example, by adjusting the final heat treatment to a low temperature or a short time, the small-diameter ratio of the Si phase tends to increase.

[0085] As mentioned above, the above is an example of the manufacture of Al bonding wire, which is a wire rod, as a representative example of Al bonding wire or Al bonding ribbon. The same procedure can also be used to manufacture Al bonding ribbon, which is a strip material. The temperature and time of the heat treatment can be approximately the same as those described above. Furthermore, when manufacturing Al bonding ribbon by rolling, the die area reduction rate can be adjusted by replacing it with the rolling reduction rate.

[0086] [Semiconductor Devices] A semiconductor device can be manufactured by connecting electrodes on a semiconductor chip to external electrodes on a lead frame or substrate using the Al bonding wire or Al bonding ribbon of the present invention. That is, the semiconductor device of the present invention includes the Al bonding wire or Al bonding ribbon of the present invention. As mentioned above, wedge bonding is used for both the first bonding with the electrode on the semiconductor chip and the second bonding with the electrode on the lead frame or substrate.

[0087] In one embodiment, the semiconductor device of the present invention comprises a circuit board, a semiconductor chip, and an Al bonding wire or Al bonding ribbon for electrically connecting the circuit board and the semiconductor chip, and is characterized in that the Al bonding wire or Al bonding ribbon is the Al bonding wire or Al bonding ribbon of the present invention.

[0088] In the semiconductor device of the present invention, the circuit board and semiconductor chip are not particularly limited, and known circuit boards and semiconductor chips that can be used to configure a semiconductor device may be used. Alternatively, a lead frame may be used instead of the circuit board. For example, as in the semiconductor device described in JP 2020-150116 A, the semiconductor device may be configured to include a lead frame and a semiconductor chip mounted on the lead frame.

[0089] Examples of semiconductor devices include various semiconductor devices used in electrical appliances (e.g., computers, mobile phones, digital cameras, televisions, air conditioners, solar power generation systems, etc.) and vehicles (e.g., motorcycles, automobiles, trains, ships, aircraft, etc.), and among these, power semiconductor devices are preferred. [Example]

[0090] The present invention will be specifically described below with reference to examples, although the present invention is not limited to the examples shown below.

[0091] (sample) The sample preparation method will be described. The raw material Al had a purity of 4N (99.99% by mass or higher), with the remainder consisting of inevitable impurities. The alloying elements Si, the first element group (Sr, Na, Eu, P), the second element group (Ni, Ti, Fe, Mg), and other elements (Mn, Zn) had a purity of 99.99% by mass or higher, with the remainder consisting of inevitable impurities. The Al alloy used for the Al bonding wire or Al bonding ribbon was produced by loading the Al raw material and the raw materials of the alloying elements into a melting crucible and using a continuous casting furnace to produce a cast product. The atmosphere inside the furnace during melting was an Ar atmosphere, and the maximum temperature of the molten metal during melting was 800°C or higher but lower than 1050°C. The cooling method after melting was air cooling (cooling in the air) or water cooling (cooling in water).

[0092] A cylindrical ingot with a diameter of 6 mm was obtained by melting. The ingot was then subjected to solution treatment and homogenization treatment, followed by wire drawing using a die and intermediate heat treatment to produce a Φ300 μm Al bonding wire. Furthermore, using the Φ300 μm Al bonding wire as the starting material, an Al bonding ribbon with a thickness of 100 μm and a width of 600 μm was produced by two-stage rolling. The temperature range of the solution treatment was 500°C or higher but lower than 550°C, and the time was 2 hours or higher but lower than 4 hours. After the solution treatment, a homogenization treatment was performed continuously during cooling. The temperature range of the homogenization treatment was 250°C or higher but lower than 350°C, and the time was 2 hours or higher but lower than 5 hours. The cooling method after the homogenization treatment was air cooling in the atmosphere.

[0093] The number of intermediate heat treatments ranged from three to four. The wire diameter after intermediate heat treatment relative to the final wire diameter was 6.5 to 7.0 times for the first intermediate annealing, 4.0 to 5.0 times for the second, and 2.0 to 3.0 times for the third. When intermediate annealing was performed four times, the wire diameter was 7.5 to 8.5 times the final wire diameter. The temperature range for the first and second intermediate heat treatments was 300°C or higher but lower than 370°C, and the time was 1 hour or higher but lower than 3 hours, while the temperature range for the third and fourth intermediate heat treatments was 250°C or higher but lower than 400°C, and the time was 2 hours or higher but lower than 40 hours.

[0094] A commercially available lubricant was used during wire drawing, and the wire area reduction rate per die during wire drawing was 10.0% or more and less than 30.0%. The wire area reduction rate was adjusted according to the wire diameter of the die. The temperature range of the final heat treatment was 200°C or more and less than 350°C, and the time of the final heat treatment was 2 hours or more and less than 20 hours. The temperature of the final heat treatment (Tc) was selected in the temperature range 50 to 100°C lower than the temperature (Tm) of the third or fourth intermediate annealing described above.

[0095] In some examples, the wire was drawn using a die with a die angle of 14° or more and less than 18°.

[0096] (Method for measuring element content) The concentration analysis of elements contained in the Al bonding wire or Al bonding ribbon was performed using an ICP-OES (Inductively Coupled Plasma-Optical Emission Spectrometer) ("PS3520UVDDII" manufactured by Hitachi High-Tech Science Corporation) or an ICP-MS (Inductively Coupled Plasma-Mass Spectrometer) ("Agilent 7700x ICP-MS" manufactured by Agilent Technologies, Inc.).

[0097] (X-ray CT analysis) The 3D X-ray microscope "Xradia 520 Versa" (ZEISS) was used. The X-ray voltage was 40 kV, the X-ray output was 3 W, the wavelength-limiting filter was LE1, and the magnification lens was 4x. Under these conditions, transmission images with transmittances of 30% to 70% were obtained. To observe minute voids, the pixel size was set to approximately 0.7 μm, and the measurement field of view was measured over an area of ​​approximately 700 μm × 700 μm × 700 μm. The sample length was approximately 10 mm, and the end of the sample was fixed. CT projection images were obtained by passing X-rays through the sample at a rotation angle of 360° around the sample axis. In this measurement, CT projection images were acquired at intervals of 0.225° per image over a 360° rotation angle (1601 images in total). Each CT projection image was observed with a 30-second exposure time. After obtaining CT projection images at all angles, reconstruction processing, such as center shifting, was performed to obtain 3D image data. Additionally, beam hardening was performed as necessary.

[0098] Next, the 3D image data was analyzed using image analysis software. The image analysis software used was Avizo Inspection. The 3D image data reflected the density differences in the observation area as a three-dimensional distribution of brightness values. Relatively high-density wire or ribbon regions (material) exhibited high brightness, while low-density voids and external space exhibited low density. First, a brightness histogram (Figure 3) was created using the measurement data, which plotted the brightness values ​​and their frequency in each voxel of the 3D image data. The brightness value at the midpoint of the two peak positions corresponding to voids or external space and material was set as the threshold. Image data areas showing brightness lower than the threshold were identified as voids or external space (the environment surrounding the sample). Next, the identified voids and external space were selected and removed from the image data area. This allowed us to identify only the image data of voids within the material within the measurement area of ​​the sample. Next, using a similar process, image data with brightness higher than the threshold was identified as image data related to material.

[0099] The image data of the voids and material thus obtained were analyzed. The method for calculating the void gradient ratio will now be described in detail. Using image analysis software, the total volume Vs of voids with a sphere-equivalent diameter of 1 μm or more and less than 10 μm and the volume of material (Vm) were calculated. The ratio (Vk / (Vk+Vg)) obtained by dividing Vk by the total volume (Vk+Vg) of voids with a sphere-equivalent diameter of 1 μm or more and less than 10 μm and material was determined as the void volume ratio. According to this method, the void volume ratios Rd and Rf of the core and surface portions were calculated, and the void gradient ratio (Rf / Rd) was calculated.

[0100] (Small diameter ratio of Si phase [Ns / Nc×100(%)]) The small diameter ratio of the Si phase [Ns / Nc × 100 (%)] in the L cross section was measured using a SEM-EDS-EBSD system, combining information on the Al and Si concentrations obtained by SEM-EDS with information on the crystal orientation obtained by EBSD. In detail, the measurement was carried out according to the following steps (1) to (3). (1) In the measurement area where the L-section of the Al bonding wire or Al bonding ribbon was used as the inspection surface, the Al and Si concentrations were measured using EDS and the crystal orientation was measured using EBSD simultaneously. (2) Using the Chi Scan function of the EBSD analysis software, Al and Si were separated and extracted. Specifically, Al and Si were separated and identified by setting a tolerance equivalent to the Si threshold from the Si EDS measurement results. The Al and Si crystal information in the material file was used for crystal orientation analysis. Here, the tolerance condition was mainly set to 30%, and was adjusted as necessary. (3) The crystal orientation of the region identified as Si phase was analyzed, and if the misorientation between measurement points was 15° or more, it was determined to be a grain boundary, and the circle-equivalent diameter of each crystal grain was calculated. The number of crystal grains identified as Si phase was tallied to determine the total number of Si phase particles, Nc. Here, Si phases with a circle-equivalent diameter of 0.5 μm or more were targeted. Considering the analytical accuracy of current EDS and EBSD analysis equipment, fine particles less than 0.5 μm were excluded. Next, the number of Si phases, Ns, with a circle-equivalent diameter in the range of 0.5 μm to 0.8 μm was tallied. The ratio of Ns to Nc [Ns / Nc × 100 (%)] (the small diameter ratio of Si phases) was calculated.

[0101] The small diameter ratio of the Si phase was determined as the average value (arithmetic mean) of the values ​​obtained for the three measurement regions by the above procedures (1) to (3).

[0102] (Evaluation method for Al bonding wire or Al bonding ribbon) The evaluation method for Al bonding wire will be explained below. The wire diameter of the Al bonding wire used for the evaluation was Φ300 μm. The semiconductor chip used was made of Si, and the electrodes on the semiconductor chip were made of a 4 μm thick film of an alloy with a composition of Al-0.5% Cu. The substrate used was an Al alloy with a 5 μm thick Ni film. A commercially available wire bonder (manufactured by Ultrasonic Industries Co., Ltd.) was used to bond the Al bonding wire, and wedge bonding was used for both the first bonding (bonding to the above electrodes on the semiconductor chip) and the second bonding (bonding to the above substrate). The Al bonding ribbon was bonded using a Hesse fully automatic bonder "BJ955" equipped with a ribbon bond head.

[0103] (Method for evaluating reliability of high-speed temperature cycles) A commercially available high-speed thermal shock tester was used for the high-speed temperature cycle test (high-speed TCT). In the high-speed TCT, hot air is blown onto the sample to rapidly heat it. The sample used for the high-speed TCT had a semiconductor chip mounted on a substrate, and electrodes on the semiconductor chip and those on the substrate were connected with Al bonding wire or Al bonding ribbon. The sample placed in the sample chamber of the high-speed thermal shock tester was subjected to repeated thermal loads, consisting of heating and cooling cycles. The minimum temperature during cooling was -40°C, and the maximum temperature during heating was 185°C. The heating time, including the heating time, was 20 seconds, and the cooling time, including the cooling time, was 40 seconds. After 10,000 cycles, the sample was removed and the shear strength of the first joint was tested. The shear strength (shear strength) of the first joint used to evaluate the high-speed temperature cycle reliability was the average shear strength of 10 randomly selected first joints. The ratio (percentage) of the average shear strength after high-speed TCT to the average shear strength before the test was taken as the strength retention rate. The higher this strength retention rate, the better the reliability of the joint. A strength retention rate of 70% or more was judged to be excellent and given a rating of "3," 60% or more but less than 70% was judged to be excellent and given a rating of "2," 50% or more but less than 60% was judged to require improvement and given a rating of "1," and less than 50% was judged to have practical problems and given a rating of "0." "3" and "2" were judged to be pass, while "1" and "0" were judged to be fail. The evaluation results are shown in the "High-speed temperature cycle reliability 10,000 cycles" column in the table.

[0104] (13,000 cycles of high-speed temperature cycling reliability evaluation) The high-speed temperature cycle test described above was conducted up to 13,000 cycles, with a lower limit of -40°C and an upper limit of 185°C. A shear test was also conducted on the first bond. The shear strength of the first bond used to evaluate high-speed temperature cycle reliability was calculated by averaging the shear strength of five randomly selected first bond locations. The bond strength retention rate, which is the ratio of shear strength after the temperature cycle test to the value before the test, was evaluated. A bond strength retention rate of less than 50% was deemed problematic for practical use and rated "0." A bond strength retention rate of 50% to less than 60% was deemed necessary and rated "1." A bond strength retention rate of 60% to less than 70% was deemed excellent and rated "2." A bond strength retention rate of 70% or greater was deemed particularly excellent and rated "3." The evaluation results are listed in the "High-speed temperature cycle reliability 13,000 cycles" column in the table.

[0105] (Method for evaluating the shear strength ratio of the first joint) This section explains the evaluation method for the shear strength ratio of the first bond. The bonding conditions were based on the standard conditions for Al bonding wire, and the ultrasonic output and load were set slightly higher to ensure a sufficient bonding area. Shear strength was evaluated using a shear test to measure the shear strength of the first bond. Ten first bonded locations were subjected to shear tests on the first bonded locations, and the average shear strength (SH) was measured. Shear strength was measured using a commercially available microshear strength tester (Nordson 4000-PLUS). The shear rate was 200 μm / s, and the shear tool height was 10 μm from the electrode surface. Shear strength measurements were performed by fixing the substrate bonded with Al bonding wire or Al bonding ribbon in a jig. Tensile tests were also performed on five samples to measure the average 0.2% proof stress (PS). The tensile tester used was the RTF-1225 (A&D), and the tensile speed was 10 mm / min. The ratio (SH / PS) obtained by dividing the shear strength (SH) by the 0.2% proof stress (PS) from the tensile test was used as the shear strength multiplier. If the shear strength multiplier was 5.2 or higher, the bond was judged to be excellent and rated as "3," if it was 4.5 or higher but less than 5.2, it was judged to have no practical problems and rated as "2," if it was 3.5 or higher but less than 4.5, it was judged to require improvement and rated as "1," and if it was less than 3.5, it was judged to have practical problems and rated as "0." The evaluation results are shown in the "Shear Strength Multiplier" column in the table.

[0106] (Method for evaluating wire breakage during processing) The method for evaluating wire breakage during processing is explained below. Wire drawing was performed with diameters ranging from 6 mm to 0.3 mm, and the number of wire breakages was confirmed. The wire drawing conditions, such as feed speed and area reduction rate, were selected from the conditions described above, and the appropriate manufacturing conditions were adjusted and changed for each wire. The drawn Al bonding wire lengths ranged from 100 to 200 m, and the number of wire breakages was calculated per 100 m. If the number of breakages was zero, it was judged to be good and given a rating of "3." If it was one breakage, it was judged that it could be addressed by improving the manufacturing conditions and given a rating of "2." If it was two to four breakages, it was considered a problem of reduced productivity and given a rating of "1." If it was five or more breakages, it was judged that it was difficult to use in practice and given a rating of "0." The evaluation results are shown in the "Breakage during processing" column in the table.

[0107] (Evaluation method for surface scratches and scrapes) The surface quality of the Al bonding wire or Al bonding ribbon was evaluated, focusing on scratches and abrasions. The Al bonding wire diameter was 0.3 mm. The Al bonding ribbon was 100 μm thick and 600 μm wide. Three measurement areas were randomly selected at intervals of at least 1 m along the central axis of the Al bonding wire or Al bonding ribbon. Three approximately 2 cm lengths were taken from each of the three areas, for a total of nine samples. The surfaces were observed using an SEM at magnifications ranging from 50 to 500x. Scratches longer than 50 μm and abrasions longer than 30 μm were considered defective. The number of scratches or abrasions was counted, and a rating of "3" was given for a good pass; one to two were considered acceptable for practical use; three to seven were considered poor for surface quality; and eight or more were considered unsuitable for practical use and rated "0." The evaluation results are listed in the "Surface Quality" column in the table.

[0108] The evaluation results of the Examples and Comparative Examples are shown in Tables 1 to 4. Examples 1 to 30 and Comparative Examples 1 to 6 in Tables 1 to 3 are results for Al bonding wires, and Examples B1 to B3 and Comparative Example B1 in Table 4 are results for Al bonding ribbons.

[0109] [Table 1]

[0110] [Table 2]

[0111] [Table 3]

[0112] [Table 4] [Explanation of symbols]

[0113] 1. Al bonding wire 10 center axis 11 L cross section 2. Al bonding ribbon 20 center axis 21 L cross section 51 Surface of Al bonding wire 52 Middle part of Al bonding wire 53 Core of Al bonding wire 61 Surface of Al bonding ribbon 62 The middle part of the Al bonding ribbon 63 Core of Al bonding ribbon

Claims

1. An Al bonding wire or Al bonding ribbon containing 3.0 mass% or more and 20.0 mass% or less of Si, A closed curve representing the periphery of the cross section perpendicular to the central axis of the Al bonding wire or Al bonding ribbon is defined as S. ０ , S ０ A closed curve that is 2 / 3 times similar to １ , S ０ A closed curve that is 1 / 3 times similar to ２ As such, S ０ , S １ and S ２ is arranged so that its center of gravity coincides with the center of gravity of the cross section perpendicular to the central axis of the Al bonding wire or Al bonding ribbon, and S ０ and S １ The area surrounded by the surface portion is called S ２ The area surrounded by is the core, An Al bonding wire or Al bonding ribbon in which the ratio (Rf / Rd) is 0.005 or more and 0.50 or less, where Rd is the volume fraction of voids in the core portion having a sphere-equivalent diameter of 1 μm or more and less than 10 μm, and Rf is the volume fraction of voids in the surface portion having a sphere-equivalent diameter of 1 μm or more and less than 10 μm, as measured by X-ray CT (Computed Topography) analysis.

2. 2. An Al bonding wire or Al bonding ribbon as described in claim 1, wherein Ns is the number of Si phases having a circle equivalent diameter of 0.5 μm or more and 0.8 μm or less in an L cross section (a cross section in the central axis direction including the central axis), and Nc is the number of Si phases having a circle equivalent diameter of 0.5 μm or more in an L cross section, and the ratio of Ns to Nc [Ns / Nc × 100 (%)] is 30% or more and 95% or less.

3. 2. The Al bonding wire or Al bonding ribbon according to claim 1, further containing one or more of Sr, Na, Eu, and P in a total amount of 10 mass ppm or more and 800 mass ppm or less.

4. The Al bonding wire or Al bonding ribbon according to claim 1, further containing one or more of Ni, Ti, Fe, and Mg in a total amount of 100 mass ppm or more and 1500 mass ppm or less.

5. The Al bonding wire or Al bonding ribbon according to claim 3, further containing one or more of Ni, Ti, Fe, and Mg in a total amount of 100 mass ppm or more and 1500 mass ppm or less.

6. The Al bonding wire or Al bonding ribbon according to any one of claims 1 to 5, wherein the total concentration of elements other than Al, Si, Sr, Na, Eu, P, Ni, Ti, Fe, and Mg in the Al bonding wire or Al bonding ribbon is 0.5 mass% or less.

7. The Al bonding wire or Al bonding ribbon according to claim 2, wherein the circle equivalent diameter and the number of Si phases are values ​​measured using a SEM-EDS-EBSD device.

8. 2. The Al bonding wire or Al bonding ribbon according to claim 1, which is for use in a semiconductor device.

9. A semiconductor device comprising the Al bonding wire or Al bonding ribbon according to claim 1.

Images