Bonding wire for semiconductor devices

Abstract

Provided is a novel Cu bonding wire for semiconductor devices with which it is possible to obtain excellent shape stability of a second bonding portion. The bonding wire for semiconductor devices comprises a core material made of Cu or a Cu alloy, and a coating layer which is formed on the surface of the core material and includes a conductive metal other than Cu, the bonding wire characterized in that: the average size in a wire circumferential direction of crystal grains according to an analysis of the surface of the wire by the electron backscatter diffraction (EBSD) method is 35-140 nm; in a concentration profile in the depth direction of the wire according to a measurement performed by the Auger electron spectroscopy (AES), three or more elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag are included in a region (hereafter referred to as "region d0-10") of a depth up to 10 nm inclusive from the surface; and the following concentration conditions (i) and (ii) are satisfied.　(i) With respect to at least three elements among the three or more elements included in the region d0-10, the average concentration of each element in the region d0-10 is more than or equal to 5 atom %; and (ii) with respect to all the elements among the three or more elements included in the region d0-10, the average concentration of each element in the region d0-10 is less than or equal to 80 atom %.

Description

Bonding wire for semiconductor devices

[0001] The present invention relates to a bonding wire for a semiconductor device, and further to a semiconductor device including the bonding wire.

[0002] In semiconductor devices, electrodes formed on a semiconductor chip are connected to electrodes on a lead frame or a substrate by bonding wires. The bonding wire connection process is completed by first bonding to an electrode on the semiconductor chip using a cylindrical bonding tool (capillary) through which the bonding wire is passed, then forming a loop and then second bonding the wire portion to an external electrode such as an electrode on a lead frame or a substrate. The first bonding is performed by heating and melting the tip of the wire portion (hereinafter also referred to as the "tail") extending from the capillary with arc heat input, forming a free air ball (FAB; hereinafter simply referred to as the "ball") by surface tension, and then crimping the ball portion to an electrode on the semiconductor chip (hereinafter also referred to as "ball bonding"). The second bonding is performed by crimping the wire portion to an external electrode by applying ultrasonic waves and a load from the capillary to the wire portion (hereinafter also referred to as "wedge bonding") without forming a ball. After the connection process, the joints are sealed with a sealing resin to obtain a semiconductor device.

[0003] Until now, gold (Au) has been the mainstream material for bonding wire, but copper (Cu) is increasingly being used instead, primarily for LSI applications (see, for example, Patent Documents 1 to 3). Furthermore, with the recent spread of electric and hybrid vehicles, it is expected that Cu, which has high thermal conductivity and fusing current, will be used in automotive devices and also in power devices (power semiconductor devices) for high-power equipment such as air conditioners and solar power generation systems. Cu is therefore expected to be a highly efficient and reliable alternative.

[0004] Cu has the disadvantage of being more easily oxidized than Au, and as a method of preventing the surface oxidation of Cu bonding wire, a structure in which the surface of the Cu core material is coated with a metal such as Pd has been proposed (Patent Document 4). Also, a Pd-coated Cu bonding wire has been proposed in which the surface of the Cu core material is coated with Pd and Pd and Pt are further added to the Cu core material to improve the bonding reliability of the first joint (Patent Document 5).

[0005] Japanese Patent Application Laid-Open No. 61-48543 Japanese Patent Application Laid-Open No. 2018-503743 International Publication No. 2017 / 221770 Japanese Patent Application Laid-Open No. 2005-167020 International Publication No. 2017 / 013796

[0006] As described above, in the second bonding process, ultrasonic waves and a load are applied from the capillary to the wire portion to crimp and bond the wire onto the external electrode. More specifically, the second bonding process consists of a stitch bonding process, in which the capillary tip presses the wire to bond it to the external electrode, and a tail bonding process, which is a temporary bonding process for forming a tail in preparation for forming a fabrication bond (FAB) in the next process. The temporary wire bond portion of the tail bonding process is formed to correspond to the edge of the wire supply opening at the capillary tip, and once a certain length of tail has been secured, the temporary wire bond portion is peeled off together with the tail. By peeling the temporary wire bond portion of the tail bonding process from the wire crimp portion deformed by the stitch bonding, the second bond portion formed on the external electrode has a fishtail shape (like a fish's tail fin) (see FIG. 1 ; the second bond portion is designated by the reference numeral 10). In this regard, as semiconductor devices become smaller and have more pins, and external electrodes are being made finer in pitch, in order to perform stable second bonding on small-area external electrodes arranged at a narrow pitch, and in order to achieve good continuous bonding, including stabilizing the size of the FAB formed in the next process, further improvements are required in the shape stability of the second bonding part, such as the symmetry and dimensional stability of the fishtail shape mentioned above, for the bonding wires used.

[0007] Furthermore, automotive devices and power devices tend to be exposed to higher temperatures during operation than general electronic devices, and the bonding wires used are required to exhibit good bonding reliability in harsh high-temperature environments. In this regard, it has been confirmed that with conventional Cu bonding wires having a Pd coating layer, the Pd coating layer partially peels off during the wire connection process, exposing the Cu core material, and the contact area between the coated Pd portion and the exposed Cu portion is exposed to an environment containing oxygen, water vapor, and sulfur compound-based outgassing generated from the sealing resin in a high-temperature environment, causing local corrosion of the Cu, i.e., galvanic corrosion, and resulting in insufficient bonding reliability at the second bond.

[0008] First, the present invention provides a novel Cu bonding wire that provides good shape stability of the second bonded portion.

[0009] Secondly, the present invention provides a novel Cu bonding wire that provides good shape stability of the second bonded portion and good bonding reliability of the second bonded portion even in a harsh high-temperature environment.

[0010] As a result of extensive research into the above problems, the present inventors have found that the above problems can be solved by providing the following configuration, and have completed the present invention.

[0011] That is, the present invention includes the following: [1] A bonding wire for a semiconductor device, comprising a core material made of Cu or a Cu alloy and a coating layer containing a conductive metal other than Cu formed on the surface of the core material, wherein the average size of crystal grains in the circumferential direction of the wire obtained by analyzing the surface of the wire by electron backscatter diffraction (EBSD) is 35 nm or more and 140 nm or less, and in a concentration profile in the depth direction of the wire obtained by measurement by Auger electron spectroscopy (AES), a region from the surface to a depth of 10 nm (hereinafter referred to as "region d") is ０－１０ ") contains three or more elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag, and satisfies the concentration conditions of (i) and (ii) below. (i) Region d ０－１０For at least three elements among the three or more elements included in the region d ０－１０ (ii) Region d: the average concentration of each element in the region is 5 atomic % or more ０－１０ For all elements among the three or more elements included in the region d ０－１０ [2] The bonding wire according to [1], wherein the average concentration of each element in the coating layer is 80 atomic % or less. [3] The bonding wire according to [1] or [2], wherein, as a result of measuring the crystal orientation of the wire surface by an EBSD method, the proportion of <111> crystal orientations in the longitudinal direction of the wire, which have an angular difference of 15 degrees or less with respect to the longitudinal direction of the wire, is 30% to 95%. [4] The bonding wire according to any of [1] to [3], wherein the concentration profile in the depth direction of the wire is obtained by measuring by AES under the following <conditions> while digging down from the surface of the wire in the depth direction by Ar sputtering. <Conditions> The center of the wire width is positioned to be the center of the width of the measurement surface, and the width of the measurement surface is 5% to 15% of the wire diameter, and the length of the measurement surface is 5 times the width of the measurement surface. [5] A bonding wire according to any one of [1] to [4], which contains one or more elements selected from the group consisting of B, P, In, and Mg (hereinafter referred to as "first additional element"), and the total concentration of the first additional element with respect to the entire wire is 1 ppm by mass to 100 ppm by mass. [6] A bonding wire according to any one of [1] to [5], which contains one or more elements selected from the group consisting of Se, Te, As, and Sb (hereinafter referred to as "second additional element"), and the total concentration of the second additional element with respect to the entire wire is 1 ppm by mass to 100 ppm by mass. [7] The bonding wire according to any one of [1] to [6], which contains one or more elements (hereinafter referred to as "third additional elements") selected from the group consisting of Ga, Ge, and Ag, and in which the total concentration of the third additional elements relative to the entire wire is 0.011 mass% or more and 1.5 mass% or less. [8] A semiconductor device including the bonding wire according to any one of [1] to [7].

[0012] According to the present invention, a novel Cu bonding wire can be provided that provides good shape stability of the second bonded portion.

[0013] The present invention also provides a novel Cu bonding wire that provides good shape stability of the second joint and good bonding reliability of the second joint even in a harsh high-temperature environment.

[0014] Fig. 1 is a schematic diagram for explaining a second bond. Fig. 2 is a schematic diagram for explaining the position and dimensions of the measurement surface when measuring the width of crystal grains on the surface of a wire by EBSD and when performing composition analysis by AES. Fig. 3 is a schematic diagram of a concentration profile in the depth direction of a wire according to one embodiment of the present invention obtained by measuring the wire by AES. Fig. 4 is a schematic diagram of a concentration profile in the depth direction of a wire according to one embodiment of the present invention obtained by measuring the wire by AES. Fig. 5 is a schematic diagram of a concentration profile in the depth direction of a wire according to one embodiment of the present invention obtained by measuring the wire by AES.

[0015] The present invention will be described in detail below with reference to preferred embodiments. While the description may refer to drawings, the drawings merely show the shapes, sizes, and arrangements 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 and their equivalents.

[0016] [Bonding Wire for Semiconductor Device] The bonding wire for semiconductor device of the present invention (hereinafter simply referred to as "wire of the present invention" or "wire") comprises a core material made of Cu or a Cu alloy, and a coating layer containing a conductive metal other than Cu formed on the surface of the core material, wherein the average size of crystal grains in the circumferential direction of the wire obtained by analyzing the surface of the wire by electron backscatter diffraction (EBSD) is 35 nm or more and 140 nm or less, and a concentration profile in the depth direction of the wire obtained by measuring by Auger electron spectroscopy (AES) is 35 nm or more and 140 nm or less. ０－１０"). The region d contains three or more elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag, and is characterized by satisfying the following concentration conditions (i) and (ii). (i) Region d ０－１０ For at least three elements among the three or more elements included in the region d ０－１０ (ii) Region d: the average concentration of each element in the region is 5 atomic % or more ０－１０ For all elements among the three or more elements included in the region d ０－１０ The average concentration of each element in the

[0017] As described above, the second joint formed on the external electrode has a fishtail shape (fish tail fin shape). FIG. 1 shows a schematic diagram of the second joint formed on the external electrode as viewed from directly above in a direction perpendicular to the main surface of the external electrode. In FIG. 1 , a fishtail-shaped second joint 10 is formed at the right end of the wire 1. To stably perform second joints on small-area external electrodes arranged at a narrow pitch, and to achieve good continuous jointability, including stabilizing the size of the FAB formed in the next process, it is preferable that the fishtail shape of the second joint exhibits good symmetry. In detail, in FIG. 1 , the axis of the wire is indicated by a dashed line X extending in the left-right direction, and it is preferable that the fishtail shape of the second joint exhibits good symmetry with respect to the axis of the wire. From the same viewpoint as above, it is preferable that the fishtail shape of the second joint exhibits good dimensional stability. Specifically, when performing multiple second bondings in succession, it is preferable to realize a fishtail shape with little variation in the deformation length in the axial direction of the wire (the dimension in the left-right direction of second bonding portion 10 in FIG. 1 ) and the deformation width in the direction perpendicular to the wire axis (the dimension 10w in the up-down direction of second bonding portion 10 in FIG. 1 ). In particular, when performing stable second bonding on small-area external electrodes provided at a narrow pitch, it is particularly preferable to realize a fishtail shape with little variation in the deformation width in the direction perpendicular to the wire axis.

[0018] The inventors have worked diligently to improve the shape stability of the second bond, such as the symmetry and dimensional stability of the fishtail shape. As a result, they have found that a bonding wire comprising a core material made of Cu or a Cu alloy and a coating layer containing a conductive metal other than Cu formed on the surface of the core material, the average size of the crystal grains in the circumferential direction of the wire obtained by analyzing the surface of the wire using an EBSD method is 35 nm to 140 nm, and the concentration profile in the depth direction of the wire obtained by measuring using AES contains three or more elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag in a region from the surface to a depth of 10 nm, and satisfies the above concentration conditions (i) and (ii). It has been found that a bonding wire having a coating layer with the above specific configuration can provide good shape stability of the second bond, even when exposed to a harsh high-temperature environment exceeding 175 ° C. As described above, the present invention significantly contributes to miniaturization and multi-pin count of semiconductor devices, and furthermore, significantly contributes to the practical application and promotion of Cu bonding wire in in-vehicle devices and power devices.

[0019] <Core Material Made of Cu or Cu Alloy> The wire of the present invention includes a core material made of Cu or a Cu alloy (hereinafter, also simply referred to as "Cu core material").

[0020] The Cu core material is not particularly limited as long as it is made of Cu or a Cu alloy, and a known Cu core material that constitutes a conventional Pd-coated Cu wire known as a bonding wire for semiconductor devices may be used.

[0021] In the present invention, the concentration of Cu in the Cu core material can be, for example, 97 atomic % or more, 97.5 atomic % or more, 98 atomic % or more, 98.5 atomic % or more, 99 atomic % or more, 99.5 atomic % or more, 99.8 atomic % or more, 99.9 atomic % or more, or 99.99 atomic % or more at the center (axial core portion) of the Cu core material.

[0022] The Cu core material may contain, for example, one or more dopants selected from the first additional element, the second additional element, and the third additional element described below. The preferred contents of these dopants are as described below.

[0023] In a preferred embodiment, the Cu core material is composed of Cu and inevitable impurities. In another preferred embodiment, the Cu core material is composed of Cu, one or more elements selected from the first additional element, the second additional element, and the third additional element described below, and inevitable impurities. Note that the term "unavoidable impurities" used in reference to the Cu core material also includes elements constituting the coating layer described below.

[0024] <Coating Layer> The wire of the present invention includes a coating layer (hereinafter also simply referred to as "coating layer") that contains a conductive metal other than Cu and is formed on the surface of a Cu core material.

[0025] From the viewpoint of providing good shape stability of the second welded portion and also from the viewpoint of providing good joining reliability of the second welded portion even under a severe high-temperature environment, it is important that the coating layer of the wire of the present invention satisfy both of the following conditions (1) and (2): (1) The average size of the crystal grains in the circumferential direction of the wire, obtained by analyzing the surface of the wire by the EBSD method, is 35 nm or more and 140 nm or less; (2) In the concentration profile in the depth direction of the wire, measured by AES, the region from the surface to a depth of 10 nm ("region d ０－１０ ") contains three or more elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag, and satisfies the following concentration conditions (i) and (ii). (i) Region d ０－１０ For at least three elements among the three or more elements included in the region d ０－１０ (ii) Region d: the average concentration of each element in the region is 5 atomic % or more ０－１０ For all elements among the three or more elements included in the region d ０－１０ The average concentration of each element in the

[0026] Condition (1) Condition (1) relates to the average size of crystal grains in the circumferential direction of the wire (the "grain width"), which is obtained by analyzing the surface of the wire by the EBSD method.

[0027] By including a coating layer that satisfies condition (1) in combination with condition (2), the wire of the present invention can provide good shape stability of the second joint. Furthermore, it can provide good joint reliability of the second joint even in harsh high-temperature environments. The crystal grains on the wire surface form a structure that is elongated in the wire longitudinal direction. Through research by the inventors, we have found that reducing the crystal grain width, which is the average size of the crystal grains in the wire circumferential direction, to a certain range is effective in improving the shape stability and joint reliability of the second joint. In comparison, we have confirmed that it is difficult to improve the shape stability and joint reliability of the second joint by controlling the average length of the crystal grains in the wire longitudinal direction or the average grain size of the crystal grains converted into a circle equivalent.

[0028] Regarding condition (1), from the viewpoint of realizing good shape stability of the second joint, particularly symmetry of the fishtail shape, the width of the crystal grains obtained by analyzing the surface of the wire of the present invention by EBSD is 140 nm or less, preferably 135 nm or less, 130 nm or less, or 125 nm or less, more preferably 120 nm or less or 110 nm or less, and even more preferably 100 nm or less, 95 nm or less, or 90 nm or less. In particular, if the width of the crystal grains is 100 nm or less, it is preferable because it is easier to realize even better symmetry of the fishtail shape. Furthermore, the inventors have confirmed that if the width of the crystal grains is within the above range, it is easier to realize excellent joining reliability of the second joint even in harsh high-temperature environments.

[0029] The lower limit of the crystal grain width is 35 nm or more, preferably 40 nm or more, 42 nm or more, or 44 nm or more, more preferably 45 nm or more, 46 nm or more, 48 nm or more, 50 nm or more, 52 nm or more, 54 nm or more, or 55 nm or more, from the viewpoint of achieving good shape stability of the second joint, particularly symmetry of the fishtail shape. In particular, when the crystal grain width is 45 nm or more, this is preferable because it makes it easier to achieve even better symmetry of the fishtail shape. Furthermore, the inventors have confirmed that when the crystal grain width is within the above range, it is easier to achieve excellent bonding reliability of the second joint even in harsh high-temperature environments.

[0030] The width of the crystal grains on the wire surface under condition (1) is obtained by analyzing the wire surface using electron backscattered diffraction (EBSD). The device used for the EBSD method is composed of a scanning electron microscope and a detector attached to it. The EBSD method is a technique in which a diffraction pattern of reflected electrons generated when a sample is irradiated with an electron beam is projected onto a detector and the diffraction pattern is analyzed to determine the crystal orientation at each measurement point. Data obtained by the EBSD method can be analyzed using analysis software (such as OIM analysis by TSL Solutions Co., Ltd.) included with the EBSD measurement device.

[0031] When measuring the width of crystal grains on the wire surface by the EBSD method, the position and dimensions of the measurement surface are determined as follows: In the following, the width of the measurement surface refers to the dimension of the measurement surface in the direction perpendicular to the wire axis (wire thickness direction, wire circumferential direction), and the length of the measurement surface refers to the dimension of the measurement surface in the direction of the wire axis (wire length direction, wire longitudinal direction).

[0032] First, the bonding wire to be measured is fixed linearly to the sample holder. Next, the wire is positioned so that the center of the wire's width in the direction perpendicular to the wire axis is the center of the width of the measurement surface, and the measurement surface is determined so that the width of the measurement surface is 20% to 40% of the wire diameter. The length of the measurement surface may be set to 2 to 5 times the width of the measurement surface. By determining the position and dimensions of the measurement surface as described above, the influence of the curvature of the wire surface can be suppressed, and the success or failure of condition (1), which is suitable for achieving good shape stability and bonding reliability of the second bond, can be accurately measured and determined. Note that it is preferable to measure at a measurement magnification in the range of 5,000 to 20,000 times and with a measurement point spacing in the range of 0.02 to 0.05 μm.

[0033] Further explanation will be given with reference to Fig. 2. Fig. 2 is a schematic plan view of a wire 1, in which the direction of the wire axis (the length direction of the wire) corresponds to the vertical direction (up and down direction) of Fig. 2, and the direction perpendicular to the wire axis (the thickness direction of the wire) corresponds to the horizontal direction (left and right direction) of Fig. 2. Fig. 2 shows the measurement surface 2 in relation to the wire 1, and the width of the measurement surface 2 is the dimension w of the measurement surface in the direction perpendicular to the wire axis. ａ and the length of the measuring surface 2 is the dimension l of the measuring surface in the direction of the wire axis. ａ The meanings of "width of the measurement surface" and "length of the measurement surface" regarding the measurement surface are the same as those for the analysis by AES under condition (2) described later. Therefore, when analyzing by the EBSD method, the measurement surface 2 is positioned so that the center of its width coincides with the dashed line X, which is the center of the wire width, and the width w of the measurement surface ａ is determined to be 20% to 40% of the wire diameter (same value as the wire width W), that is, 0.2W to 0.4W. ａ Ha, 2w ａ ≦l ａ ≦5w ａ Satisfy the relationship.

[0034] When measuring the width of crystal grains on the wire surface using the EBSD method, in order to avoid the influence of dirt, deposits, unevenness, scratches, etc. on the wire surface, it is preferable to use only crystal orientations within the measurement surface that can be identified based on a certain reliability standard, and to exclude areas where the crystal orientation cannot be measured or where the crystal orientation can be measured but the reliability of the orientation analysis is low. For example, when using OIM analysis by TSL Solutions as analysis software, it is preferable to exclude measurement points with a CI value (Confidence Index) of less than 0.1. Note that if the excluded data exceeds, for example, 30% of the total, it is likely that the measurement target has been contaminated, so it is desirable to perform the measurement again, starting with preparing the measurement sample. Here, "crystal grains on the wire surface" refers not only to crystal grains exposed on the wire surface but also to crystal grains recognized as crystal grains in EBSD measurement.

[0035] Furthermore, when measuring the width of crystal grains on the wire surface using the EBSD method, it is preferable to consider boundaries where the orientation difference between adjacent measurement points is 5 degrees or more as crystal grain boundaries and calculate the width of the crystal grains. The calculation of the width of crystal grains on the wire surface using analysis software is generally performed by (i) drawing lines in the width direction of the measurement surface (circumferential direction of the wire), determining the size of each crystal grain in the circumferential direction of the wire from the spacing of the crystal grain boundaries on the lines, and (ii) arithmetically averaging the sizes of each crystal grain in the circumferential direction of the wire to calculate the average size of the crystal grains in the circumferential direction of the wire. This is performed for multiple lines (the number of lines is preferably 10 or more, more preferably 20 or more) spaced apart from each other in the longitudinal direction of the wire, and the average value is used as the width of the crystal grains.

[0036] The width of the crystal grains on the wire surface under condition (1) is based on the results of measurements under the conditions described in the section "Crystalline analysis of wire surface by electron backscattered diffraction (EBSD) method" below.

[0037] Condition (2) Condition (2) relates to the composition of the region from the surface of the wire to a depth of 10 nm in the concentration profile in the depth direction of the wire (hereinafter simply referred to as the “depth direction concentration profile”) obtained by measurement using AES.

[0038] By including a coating layer that satisfies condition (2) in combination with condition (1), the wire of the present invention can provide good shape stability at the second joint and, further, good joining reliability at the second joint even under harsh high-temperature environments.

[0039] Although the reason for this effect is unclear, in a Cu-based bonding wire having a coating layer containing three or more elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag in a specific concentration range in a region from the surface to a depth of 10 nm, if the width of the crystal grains on the wire surface is reduced to a certain range, the synergistic effect of the alloy composition and the fine crystal grains on the surface improves the sliding properties between the capillary and the wire surface, suppresses damage to the coating layer, and achieves uniform compressive deformation, and is thought to improve the symmetry of the fishtail shape with respect to the wire axis and the dimensional stability of the fishtail shape. Furthermore, this effect suppresses damage to the coating layer, thereby suppressing the occurrence of corrosion in high-temperature environments, and is thought to improve the bonding reliability of the second joint. In this regard, in the case where the coating layer contains only Pd and the region from the surface to a depth of 10 nm contains two or less elements, or even if the region contains three or more elements, it does not contain three or more elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag. It has been confirmed that even if the width of the crystal grains is reduced to a certain range, the shape stability and bonding reliability of the second joint (particularly the shape stability of the second joint) are not improved, and it has been clarified that the above effect achieved when the width of the crystal grains is reduced to a certain range is specifically manifested in a Cu-based bonding wire having a coating layer with a specific surface composition.

[0040] In the present invention, when obtaining a concentration profile in the depth direction of the wire by AES, it is preferable to measure the depth direction so that the measurement points are 10 to 20 points in the region from the surface of the wire to a depth of 10 nm. This makes it possible to accurately measure and determine whether condition (2) is met, which is suitable for achieving good shape stability and bonding reliability of the second bond. Therefore, in a preferred embodiment, the coating layer in the wire of the present invention satisfies the above condition (2) in the concentration profile in the depth direction of the wire obtained by AES measurement so that the measurement points in the depth direction are 10 to 20 points in the region from the surface of the wire to a depth of 10 nm.

[0041] Regarding condition (2), from the viewpoint of realizing good shape stability of the second bonded portion, particularly a fishtail shape with small variations in deformation width in the direction perpendicular to the wire axis, the region from the surface of the wire to a depth of 10 nm in the concentration profile in the depth direction of the wire ("region d ０－１０ ") contains three or more elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag. ０－１０ The elements contained in the region d are two or less kinds. ０－１０ The present inventors have confirmed that even if three or more elements are contained in the second bonded portion, if the second bonded portion does not contain three or more elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag, the shape stability of the second bonded portion tends to be poor.

[0042] Regarding condition (2), in order to realize a good shape stability of the second bonded portion, particularly a fishtail shape with small variations in the deformation width in the direction perpendicular to the axis of the wire, the region d ０－１０ The three or more elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag contained in region d satisfy the following concentration conditions (i) and (ii): ０－１０ For at least three elements among the three or more elements included in the region d ０－１０ (ii) Region d: the average concentration of each element in the region is 5 atomic % or more ０－１０ For all elements among the three or more elements included in the region d ０－１０ The average concentration of each element in the

[0043] For the concentration condition (i), the area d ０－１０ For at least three elements among the three or more elements included in the region d ０－１０ The average concentration of each element in region d is 5 atomic % or more, preferably 6 atomic % or more or 8 atomic % or more, more preferably 10 atomic % or more, 12 atomic % or more, 14 atomic % or more, or 15 atomic % or more. ０－１０ For at least three elements among the three or more elements included in the region d ０－１０When the average concentration of each element in the wire is 10 atomic % or more, it is easy to realize a fishtail shape with even smaller variation in deformation width in the direction perpendicular to the wire axis, which is preferable. Furthermore, the inventors have confirmed that it is easy to realize excellent bonding reliability of the second bonded portion even in a harsh high-temperature environment.

[0044] As mentioned above, area d ０－１０ For at least three elements among the three or more elements included in the region d ０－１０ It is important that the average concentration of each element in region d is 5 atomic % or more. The following (a) to (c) can be cited as examples that satisfy the concentration condition (i), and the effects of the present invention can be obtained in any of the following examples. (a) Region d ０－１０ For all elements among the three or more elements included in the region d ０－１０ The average concentration of each element in the region is 5 atomic % or more (the preferred range of the average concentration is as described above). (b) Region d ０－１０ For three elements among the three or more elements included in region d ０－１０ The average concentration of each element in the region d is 5 atomic % or more (the preferred range of the average concentration is as described above), and the other one or two elements are in the region d ０－１０ The average concentration of each element in the region is less than 5 atomic percent. ０－１０ For four elements among the three or more elements included in region d ０－１０ The average concentration of each element in the region d is 5 atomic % or more (the preferred range of the average concentration is as described above), and the ... ０－１０ The average concentration of each element in the region is less than 5 atomic percent. ０－１０ (b) applies to a case where three to five elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag are contained, (b) applies to a case where four or five elements are contained, and (c) applies to a case where five elements are contained.

[0045] Regarding the concentration condition (ii), the region d ０－１０ For all elements among the three or more elements included in the region d ０－１０The average concentration of each element in region d is 80 atomic % or less, preferably 75 atomic % or less, 74 atomic % or less, or 72 atomic % or less, and more preferably 70 atomic % or less, 68 atomic % or less, 66 atomic % or less, or 65 atomic % or less. ０－１０ When the average concentration of each element in the wire is 70 atomic % or less, a fishtail shape with even smaller variations in deformation width in the direction perpendicular to the wire axis can be easily realized, which is preferable. Furthermore, the inventors have confirmed that it is easy to achieve excellent bonding reliability of the second bonded portion even in a harsh high-temperature environment.

[0046] Area d under condition (2) ０－１０ The composition of the wire can be confirmed and determined by performing composition analysis by AES while digging from the surface of the wire in the depth direction (toward the center of the wire) by Ar sputtering. In detail, after 1) performing composition analysis of the wire surface, 2) sputtering with Ar and 3) repeating composition analysis of the surface after sputtering, the concentration change of each element in the depth direction (center) from the surface of the wire (so-called depth direction concentration profile) can be obtained, and the composition can be confirmed and determined based on this concentration profile. In the present invention, when obtaining the depth direction concentration profile, the unit of depth is SiO ２ Here, when performing composition analysis by AES, gas components such as carbon (C), sulfur (S), oxygen (O), and nitrogen (N), non-metallic elements, etc. are not taken into consideration.

[0047] When performing 1) composition analysis of the wire surface and 3) composition analysis of the surface after sputtering, the position and dimensions of the measurement surface are determined as follows.

[0048] First, the bonding wire to be measured is fixed linearly to the sample holder. Next, the wire is positioned so that the center of the wire width in the direction perpendicular to the wire axis is the center of the width of the measurement surface, and the measurement surface is determined so that the width of the measurement surface is 5% to 15% of the wire diameter. The length of the measurement surface is set to be five times the width of the measurement surface. This will be further explained with reference to Figure 2. As mentioned above, in Figure 2, the width of the wire is indicated by the symbol W, and the center of the wire width is indicated by the dashed line X. Therefore, the measurement surface 2 is positioned so that the center of its width coincides with the dashed line X, which is the center of the wire width, and the width w of the measurement surface is determined so that the center of its width coincides with the dashed line X, which is the center of the wire width. ａ is determined to be 5% to 15% of the wire diameter (same value as the wire width W), that is, 0.05W to 0.15W. ａ Is, l ａ =5w ａ By determining the position and dimensions of the measurement surface as described above, it is possible to accurately measure whether condition (2) is met, which is suitable for achieving good shape stability and bonding reliability of the second bonded portion.

[0049] In the present invention, when obtaining a concentration profile in the depth direction of the wire by composition analysis using AES, it is preferable to fix the peaks used to detect each of the elements Pd, Pt, Au, Ni, Ag, and Cu in order to improve the accuracy of the analysis. That is, it is preferable to use the following peaks as the negative peak (minimum value) energy values ​​of each element in the Auger electron differential spectrum (hereinafter referred to as "Auger electron spectrum"): Au (2022 eV), Pd (333 eV), Pt (1969 eV), Ag (359 eV), Ni (849 eV), and Cu (922 eV) (source of energy values: data built into "MultiPak," an electron spectroscopy data analysis software manufactured by ULVAC-PHI, Inc.).

[0050] When analyzing Auger electron spectra, linear least squares (LLS) processing can be performed as needed. This allows for more accurate quantification of element concentrations. For example, when peaks of elements such as Cu and Ni have similar energy values ​​and overlap, it is preferable to perform LLS processing when separating the overlapping peak components. Furthermore, when the influence of background noise is relatively high, such as when Au or Pt is contained at low concentrations, it is preferable to perform LLS processing to reduce the background noise. Analysis of Auger electron spectra can be performed, for example, using analysis software (MultiPak) provided in an AES apparatus manufactured by ULVAC-PHI, Inc.

[0051] In the obtained concentration profile in the depth direction, the concentration C of Pd Ｐｄ (atomic %) in the region from the wire surface to a depth of 10 nm (i.e., region d ０－１０ ) C for all measurement points Ｐｄ By arithmetically averaging the values ​​of ０－１０ Similarly, the average Pd concentration C Ｐｔ (atomic %), and the region d ０－１０ C for all measurement points Ｐｔ By arithmetically averaging the values ​​of ０－１０ The average concentration of each element of Au, Ni, and Ag can be calculated in the same manner as above. ０－１０ The concentration values ​​of each element for all measurement points in the region d are arithmetically averaged. ０－１０ Then, the average concentration of each element in the region d ０－１０ It is determined that condition (2) is satisfied when the wire contains three or more elements selected from Pd, Pt, Au, Ni, and Ag and satisfies the above-mentioned concentration conditions (i) and (ii). It is also preferable to obtain concentration profiles for multiple measurement surfaces (n≧3) spaced 1 mm or more apart from each other in the wire axial direction and use the arithmetic mean value of the obtained concentration profiles.

[0052] The presence and average concentration of each element under condition (2) are based on the results of measurements under the conditions described in the section "Analysis of coating layer thickness by Auger electron spectroscopy (AES)" below.

[0053] From the viewpoint of being able to enjoy the effects of the present invention more effectively, region d ０－１０ The Pd concentration C Ｐｄ (atomic %), Pt concentration C Ｐｔ (atomic %), Au concentration C Ａｕ (atomic %), Ni concentration C Ｎｉ (atomic %), Ag concentration C Ａｇ Total C (atomic %) Ｐｄ +C Ｐｔ +C Ａｕ +C Ｎｉ +C Ａｇ The average value of is preferably 70 atomic % or more, more preferably 80 atomic % or more, 85 atomic % or more, or 90 atomic % or more, and even more preferably 92 atomic % or more, 94 atomic % or more, 95 atomic % or more, 96 atomic % or more, 98 atomic % or more, or 99 atomic % or more, and the upper limit may be 100 atomic %.

[0054] Such a total C Ｐｄ +C Ｐｔ +C Ａｕ +C Ｎｉ +C Ａｇ The average value of the concentration profile in the depth direction is ０－１０ The total C for all measurement points Ｐｄ +C Ｐｔ +C Ａｕ +C Ｎｉ +C Ａｇ The arithmetic mean of the values ​​of ０－１０ It can be found by adding up the average concentrations of the elements Pd, Pt, Au, Ni, and Ag in the above.

[0055] Other Preferred Conditions for the Coating Layer In the wire of the present invention, the coating layer more preferably satisfies one or both of the following conditions (3) and (4) in addition to satisfying the above conditions (1) and (2): (3) The thickness of the coating layer is 40 nm or more and 200 nm or less, and (4) When the crystal orientation of the wire surface is measured by the EBSD method, the proportion of <111> crystal orientations that have an angular difference of 15 degrees or less with respect to the longitudinal direction of the wire is 30% or more and 95% or less.

[0056] Condition (3) Condition (3) relates to the thickness of the coating layer.

[0057] By including a coating layer that satisfies condition (3) in addition to conditions (1) and (2), the wire of the present invention can provide even better bonding reliability of the second bonded portion even in a harsh high-temperature environment.

[0058] Regarding condition (3), the thickness of the coating layer (a calculation method based on the concentration profile in the depth direction of the wire will be described later) is preferably 40 nm or more from the viewpoint of realizing good bonding reliability of the second bond in a high-temperature environment, more preferably 42 nm or more, 44 nm or more, 45 nm or more, 46 nm or more, or 48 nm or more, more preferably 50 nm or more, 55 nm or more, 60 nm or more, 65 nm or more, or 70 nm or more. The inventors have confirmed that, in combination with conditions (1) and (2), a coating layer thickness of 50 nm or more is particularly likely to achieve exceptionally good bonding reliability of the second bond even in a harsh high-temperature environment. Furthermore, the upper limit of the thickness of the coating layer is preferably 200 nm or less from the viewpoint of realizing good bonding reliability of the second bond in a high-temperature environment, more preferably 10 nm or less, 170 nm or less, or 160 nm or less, and even more preferably 150 nm or less, 140 nm or less, 135 nm or less, or 130 nm or less. The inventors have confirmed that, in combination with conditions (1) and (2), particularly when the thickness of the coating layer is 150 nm or less, it is easy to achieve exceptionally excellent bonding reliability of the second bonding portion even in harsh high-temperature environments.

[0059] The thickness of the coating layer can be confirmed by performing composition analysis by AES while digging from the surface of the wire in the depth direction (toward the center of the wire) by Ar sputtering, as described in relation to condition (2). Specifically, after 1) performing composition analysis of the wire surface, 2) sputtering with Ar and 3) repeating composition analysis of the surface after sputtering obtain a concentration profile in the depth direction, and the thickness can be confirmed from the concentration profile. The position and dimensions of the measurement surface when performing composition analysis by AES are also as described in relation to condition (2). By determining the position and dimensions of the measurement surface as described above, the success or failure of condition (3), which is suitable for achieving good bonding reliability of the second bond, can be accurately measured.

[0060] The thickness of the coating layer can be determined from the obtained concentration profile in the depth direction. First, the boundary between the Cu core material and the coating layer is determined based on the Cu concentration. The position where the Cu concentration is 50 atomic % is determined as the boundary, and the region where the Cu concentration is 50 atomic % or more is the Cu core material, and the region where the Cu concentration is less than 50 atomic % is the coating layer. In the present invention, the boundary between the Cu core material and the coating layer does not necessarily have to be a grain boundary. Then, the concentration profile is confirmed from the wire surface toward the wire center, and the thickness can be determined as the distance from the wire surface position Z0 to the depth position Z1 where the concentration of the core material Cu first reaches 50 atomic %. In the present invention, when determining the thickness of the coating layer from the concentration profile in the depth direction, the unit of depth is SiO ２ It is also preferable to obtain concentration profiles for a plurality of measurement surfaces (n≧3) spaced apart from each other by 1 mm or more in the wire axial direction, and use the arithmetic mean value.

[0061] The thickness of the coating layer under condition (3) is based on the results of measurement under the conditions described in the section "Analysis of coating layer thickness by Auger electron spectroscopy (AES)" below.

[0062] Condition (4) Condition (4) relates to the proportion of <111> crystal orientations in the longitudinal direction of the wire, which are determined by measuring the crystal orientation of the surface of the wire using the EBSD method, that have an angular difference of 15 degrees or less with respect to the longitudinal direction of the wire (hereinafter simply referred to as the "proportion of <111> crystal orientations on the wire surface").

[0063] As mentioned above, the connection process using a bonding wire is completed by first bonding to an electrode on a semiconductor chip, then forming a loop, and then second bonding of the wire portion to an external electrode such as an electrode on a lead frame or a substrate. By including a coating layer that satisfies condition (4) in addition to conditions (1) and (2), a bonding wire that can stably form a desired loop shape can be realized.

[0064] Regarding condition (4), from the viewpoint of realizing a wire exhibiting good loop shape stability, the proportion of the <111> crystal orientation on the wire surface is preferably 30% or more, more preferably 35% or more, even more preferably 40% or more, 45% or more, 50% or more, 55% or more, or 60%, and the upper limit is preferably 95% or less, more preferably 90% or less, even more preferably 85% or less, 84% or less, 82% or less, or 80% or less. In particular, when the proportion of the <111> crystal orientation on the wire surface is in the range of 40% or more and 85% or less, a bonding wire exhibiting exceptionally good loop shape stability can be realized.

[0065] The proportion of the <111> crystal orientation on the wire surface is obtained by measuring the crystal orientation on the wire surface by the EBSD method. The equipment and analysis software used for the EBSD method are as described in relation to the above-mentioned condition (1), and the wire surface is used as the measurement surface, as well as the position and dimensions of the measurement surface. Furthermore, as described in relation to the above-mentioned condition (1), in the analysis, only the crystal orientation within the measurement surface that can be identified based on a certain reliability is used, and portions where the crystal orientation cannot be measured or portions where the crystal orientation can be measured but the reliability of the orientation analysis is low are excluded from the calculation. In the present invention, it is preferable that the proportion of the <111> crystal orientation on the wire surface is measured at multiple locations (n≧3) spaced apart from each other by 1 mm or more in the longitudinal direction of the wire, and the arithmetic average value of the measurements is used.

[0066] The proportion of the <111> crystal orientation on the wire surface tends to fall within the desired range by adjusting the composition of the coating layer, the degree of wiredrawing, heating conditions, etc. In addition to adjusting the composition of the coating layer, the conditions for increasing the proportion of the <111> crystal orientation on the wire surface can be adjusted by, for example, increasing the working rate, lowering the heating temperature, shortening the heating time, etc., provided that the composition remains the same.

[0067] The proportion of the <111> crystal orientation on the wire surface under condition (4) is based on the results of measurements under the conditions described in the section below titled "Crystalline analysis of wire surface by electron backscattered diffraction (EBSD) method."

[0068] The coating layer may contain, for example, one or more dopants selected from the first additional element, the second additional element, and the third additional element described below. The preferred contents of these dopants are as described below.

[0069] In one embodiment, the coating layer comprises three or more elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag, and inevitable impurities. In another embodiment, the coating layer comprises three or more elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag, one or more elements selected from the first additional element, the second additional element, and the third additional element described below, and inevitable impurities. Note that the term "unavoidable impurities" used in reference to the coating layer also includes the elements that make up the Cu core material described above.

[0070] As described above, the wire of the present invention including a coating layer satisfying conditions (1) and (2) can provide good shape stability of the second joint and good joining reliability of the second joint even under severe high-temperature environments. When the wire includes a coating layer that satisfies condition (3) in addition to conditions (1) and (2), a wire that exhibits particularly good joining reliability of the second joint even under severe high-temperature environments can be realized. Furthermore, when the wire includes a coating layer that satisfies condition (4) in addition to conditions (1) and (2), a wire that exhibits good loop shape stability can be realized.

[0071] The wire of the present invention can achieve the desired effects regardless of the composition in the region more than 10 nm from the surface as long as it satisfies the above conditions (1) and (2), and more preferably satisfies one or both of the above conditions (3) and (4) in addition to the above conditions (1) and (2). In particular, from the viewpoint of further enjoying the effects of the present invention, in the concentration profile in the depth direction of the wire, when the thickness of the coating layer is dt (nm), the total concentration of elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag at the position of the depth dt (nm) from the wire surface, i.e., at the boundary between the coating layer and the Cu core material, is preferably 35 atomic % or more, more preferably 40 atomic % or more, 42.5 atomic % or more, or 45 atomic % or more, and even more preferably 46 atomic % or more, 47 atomic % or more, 47.5 atomic % or more, 48 atomic % or more, 49 atomic % or more, or 49.5 atomic % or more, and the upper limit may be 50 mass %. Similarly, from the viewpoint of being able to enjoy the effects of the present invention even more effectively, in the concentration profile in the depth direction of the wire, when the thickness of the coating layer is dt (nm), at a position 0.5 dt (nm) deep from the wire surface, the total concentration of elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag is preferably 70 atomic % or more, more preferably 80 atomic % or more, 85 atomic % or more, or 90 atomic % or more, and even more preferably 92 atomic % or more, 94 atomic % or more, 95 atomic % or more, 96 atomic % or more, 98 atomic % or more, or 99 atomic % or more, and the upper limit may be 100 mass %.

[0072] From the viewpoint of being able to achieve good shape stability of the second joint, and further from the viewpoint of being able to achieve good joining reliability of the second joint even in harsh high-temperature environments, an example of a concentration profile in the depth direction of the wire according to a particularly preferred embodiment is shown below.

[0073] 3 is a schematic diagram of a concentration profile in the depth direction of a wire according to an embodiment of the present invention obtained by measuring the wire by AES. The concentration profile in the depth direction shown in FIG. 3 schematically shows the change in concentration (mol% (atomic%); vertical axis) of four elements, A element, B element, C element, and Cu, relative to the sputtering depth (nm; horizontal axis). Here, the A element, B element, and C element are each elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag. In the concentration profile in the depth direction shown in FIG. 3, from the surface of the wire (position of sputtering depth 0 nm) to a very shallow position (for example, a position at a depth of 0.3 dt, 0.25 dt, 0.2 dt, or 0.15 dt, where dt is the thickness of the coating layer), the concentration of the A element decreases and the concentrations of the B element and the C element increase. Further in the depth direction, up to a certain depth position (for example, a depth of 0.9 dt, 0.8 dt, 0.7 dt, or 0.6 dt, where dt is the thickness of the coating layer), the B element and the C element coexist in a predetermined ratio. Further in the depth direction, the concentrations of the B element and the C element decrease while the concentration of Cu increases. A wire having such a concentration profile in the depth direction may be formed, for example, by providing a Cu core with an alloy layer containing the B element and the C element in a predetermined ratio, and further providing a layer of the A element on top of that.

[0074] FIG. 4 is a schematic diagram of a concentration profile in the depth direction of a wire according to an embodiment of the present invention obtained by measuring the wire using AES. As in FIG. 3, the concentration profile in the depth direction shown in FIG. 4 also schematically shows the change in concentration (mol% (atomic%); vertical axis) of four elements, A element, B element, C element, and Cu, relative to the sputtering depth (nm; horizontal axis), where A element, B element, and C element are each selected from the group consisting of Pd, Pt, Au, Ni, and Ag. In the concentration profile in the depth direction shown in FIG. 4, from the surface of the wire (position of sputtering depth 0 nm) to a very shallow position (for example, a position at a depth of 0.3 dt, 0.25 dt, 0.2 dt, or 0.15 dt, where dt is the thickness of the coating layer), the concentration of A element decreases and the concentrations of B element and C element increase. As the depth direction progresses, the concentrations of B element and C element each reach their maximum values, and then the concentrations of B element and C element decrease and the concentration of Cu increases. The B element may exhibit a maximum concentration at a certain depth position (d1) or may exhibit a maximum value across a certain depth range (d1 to d2). The C element may exhibit a maximum concentration at a certain depth position (d3) or may exhibit a maximum value across a certain depth range (d3 to d4). Here, d1, d2, d3, and d4 satisfy the relationships d1 < d3 and d2 < d4. Furthermore, d1 may satisfy, for example, 0.05 dt < d1 < 0.9 dt or 0.1 dt < d1 < 0.8 dt, and d3 may satisfy, for example, 0.05 dt < d3 < 0.9 dt or 0.1 dt < d3 < 0.8 dt. A wire having such a depth concentration profile may be formed, for example, by providing a layer of C element on a Cu core material, providing a layer of B element thereon, and further providing a layer of A element thereon.

[0075] 5 is a schematic diagram of a concentration profile in the depth direction of a wire according to an embodiment of the present invention obtained by measuring the wire by AES. As in FIG. 3, the concentration profile in the depth direction shown in FIG. 5 also schematically shows the change in concentration (mol% (atomic%); vertical axis) of four elements, A element, B element, C element, and Cu, with respect to the sputtering depth (nm; horizontal axis), where A element, B element, and C element are each selected from the group consisting of Pd, Pt, Au, Ni, and Ag. In the concentration profile in the depth direction shown in FIG. 5, from the surface of the wire (position of sputtering depth 0 nm) to a certain depth position (for example, a position of depth 0.9 dt, 0.8 dt, 0.7 dt, or 0.6 dt, where dt is the thickness of the coating layer), the A element, B element, and C element coexist in a predetermined ratio. Further in the depth direction, the concentrations of the A element, B element, and C element decrease, while the Cu concentration increases. A wire having such a concentration profile in the depth direction may be formed, for example, by providing a Cu core material with an alloy layer containing elements A, B, and C in a predetermined ratio.

[0076] The wire of the present invention should satisfy the area d ０～１０The composition in is as explained in relation to condition (2), but from the viewpoint of realizing good shape stability of the second bonded portion and further from the viewpoint of realizing good bonding reliability of the second bonded portion even in a severe high-temperature environment, it is more preferable that the element A in the concentration profiles in the depth direction shown in Figures 3 to 5 be selected from the group consisting of Au, Pd, and Ag. From the same viewpoint, it is more preferable that the element B in the concentration profiles in the depth direction shown in Figures 3 to 5 be selected from the group consisting of Pd, Pt, Ni, and Ag, and it is more preferable that the element C be selected from the group consisting of Pd, Pt, and Ni. Therefore, in a preferred embodiment, the wire of the present invention contains, in a concentration profile in the depth direction of the wire, an element A selected from the group consisting of Au, Pd, and Ag in a region from the surface to a depth of 10 nm, and further contains at least an element B selected from the group consisting of Pd, Pt, Ni, and Ag, and an element C selected from the group consisting of Pd, Pt, and Ni (wherein the elements A, B, and C are different from one another), and the average concentration of each of these elements A, B, and C is in the range of 5 atomic % to 80 atomic %. Here, the preferred lower limit of the average concentration of each element is as described above for concentration condition (i), and the preferred upper limit of the average concentration of each element is as described above for concentration condition (ii). In particular, from the viewpoint of realizing particularly good shape stability of the second bonded portion and further from the viewpoint of realizing particularly good bonding reliability of the second bonded portion even in a severe high-temperature environment, the wire of the present invention contains at least Au, Pd, and Ni in a region from the surface to a depth of 10 nm in a concentration profile in the depth direction of the wire, and the average concentration of each of these Au, Pd, and Ni is in the range of 5 atomic % to 80 atomic %. In such an embodiment, it is preferable that the A element contains Au, the B element contains Pd or Ni, and the C element contains Ni or Pd.

[0077] 3 to 5 show the change in concentration with respect to the sputtering depth for the four elements A, B, C, and Cu, but the wire of the present invention may contain other elements (for example, elements other than A to C selected from the group consisting of Pd, Pt, Au, Ni, and Ag, the first additional element, second additional element, and third additional element described below) as long as the wire satisfies the above conditions (1) and (2). Even when the wire contains elements other than A, B, C, and Cu, the satisfaction of condition (2) can be determined by focusing on the concentration of the element selected from the group consisting of Pd, Pt, Ni, and Ag.

[0078] The wire of the present invention may further contain one or more elements ("first additional elements") selected from the group consisting of B, P, In, and Mg. When the wire of the present invention contains the first additional element, the total concentration of the first additional element relative to the entire wire is preferably 1 ppm by mass or more. This makes it possible to realize a bonding wire that provides a better crimped shape of the first bonded portion. The total concentration of the first additional element relative to the entire wire is more preferably 2 ppm by mass or more, and even more preferably 3 ppm by mass or more, 5 ppm by mass or more, 8 ppm by mass or more, 10 ppm by mass or more, 15 ppm by mass or more, or 20 ppm by mass or more. From the viewpoint of suppressing hardening of the wire and reducing chip damage during the first bonding, the total concentration of the first additional element is preferably 100 ppm by mass or less, more preferably 90 ppm by mass or less, 80 ppm by mass or less, 70 ppm by mass or less, 60 ppm by mass or less, or 50 ppm by mass or less. Therefore, in one preferred embodiment, the wire of the present invention contains a first additional element, and the total concentration of the first additional element in the entire wire is 1 mass ppm or more and 100 mass ppm or less.

[0079] When the wire of the present invention contains the first additive element, the first additive element may be contained in either the Cu core material or the coating layer, or may be contained in both. From the viewpoint of realizing a bonding wire that provides a better crimped shape of the first bonded portion, it is preferable that the first additive element be contained in the Cu core material.

[0080] The wire of the present invention may further contain one or more elements ("second additional elements") selected from the group consisting of Se, Te, As, and Sb. When the wire of the present invention contains the second additional elements, the total concentration of the second additional elements relative to the entire wire is preferably 1 ppm by mass or more. This can improve the bonding reliability of the first bonded portion in a high-temperature, high-humidity environment. The total concentration of the second additional elements relative to the entire wire is more preferably 2 ppm by mass or more, and even more preferably 3 ppm by mass or more, 5 ppm by mass or more, 8 ppm by mass or more, 10 ppm by mass or more, 15 ppm by mass or more, or 20 ppm by mass or more. From the viewpoint of realizing a good crimp shape of the first bonded portion, the total concentration of the second additional elements is preferably 100 ppm by mass or less, and more preferably 90 ppm by mass or less, 80 ppm by mass or less, 70 ppm by mass or less, 60 ppm by mass or less, or 50 ppm by mass or less. Therefore, in a preferred embodiment, the wire of the present invention contains a second additional element, and the total concentration of the second additional element in the entire wire is 1 mass ppm or more and 100 mass ppm or less.

[0081] When the wire of the present invention contains the second additive element, the second additive element may be contained in either the Cu core material or the coating layer, or may be contained in both. From the viewpoint of realizing a bonding wire that provides even better bonding reliability of the first bonded portion under high-temperature and high-humidity environments, it is preferable that the second additive element be contained in the coating layer.

[0082] The wire of the present invention may further contain one or more elements ("third additional element") selected from the group consisting of Ga, Ge, and Ag. When the wire of the present invention contains the third additional element, the total concentration of the third additional element relative to the entire wire is preferably 0.011 mass% or more. This can improve the bonding reliability of the first bonded portion in high-temperature environments. The total concentration of the third additional element relative to the entire wire is more preferably 0.015 mass% or more, and even more preferably 0.02 mass% or more, 0.025 mass% or more, 0.03 mass% or more, 0.031 mass% or more, 0.035 mass% or more, 0.04 mass% or more, 0.05 mass% or more, 0.07 mass% or more, 0.09 mass% or more, 0.1 mass% or more, 0.12 mass% or more, 0.14 mass% or more, 0.15 mass% or more, or 0.2 mass% or more. From the viewpoint of realizing a good crimped shape of the first bonded portion and good bondability at the second bonded portion, the total concentration of the third additional element is preferably 1.5 mass% or less, and more preferably 1.4 mass% or less, 1.3 mass% or less, or 1.2 mass% or less. Therefore, in a preferred embodiment, the wire of the present invention contains the third additional element, and the total concentration of the third additional element with respect to the entire wire is 0.011 mass% or more and 1.5 mass% or less.

[0083] When the wire of the present invention contains a third additional element, the third additional element may be contained in either the Cu core material or the coating layer, or may be contained in both.

[0084] The contents of the first additional element, the second additional element, and the third additional element in the wire can be measured by the method described later in [Measurement of element contents].

[0085] The diameter of the wire of the present invention is not particularly limited and may be appropriately determined depending on the specific purpose. The lower limit of the diameter may be, for example, 15 μm or more, 16 μm or more, and the upper limit of the diameter may be, for example, 80 μm or less, 70 μm or less, or 50 μm or less.

[0086] <Method for Manufacturing Wire> An example of a method for manufacturing the bonding wire for semiconductor device of the present invention will be described.

[0087] First, high-purity (4N to 6N; 99.99 to 99.9999 mass % or more) raw copper is processed into a large-diameter (approximately 3 to 7 mm) ingot by continuous casting.

[0088] When dopants such as the first, second, and third additive elements are added, the addition method can be, for example, by incorporating them into the Cu core material, incorporating them into the coating layer, depositing them on the surface of the Cu core material, or depositing them on the surface of the coating layer. A combination of these methods may also be used. The effects of the present invention can be achieved regardless of the addition method employed. In the method of incorporating a dopant into the Cu core material, a copper alloy containing the dopant at the required concentration can be used as a raw material to produce the Cu core material. When adding a dopant to the raw material Cu to obtain such a copper alloy, a high-purity dopant component can be directly added to Cu, or a mother alloy containing about 1% of the dopant component can be used. In the method of incorporating a dopant into the coating layer, the dopant can be incorporated into a Pd, Pt, Au, Ni, or Ag plating bath (in the case of wet plating) or a target material (in the case of dry plating) used to form the coating layer. In the method of depositing on the surface of a Cu core material or on the surface of a coating layer, the surface of the Cu core material or the surface of the coating layer can be used as the deposition surface, and one or more deposition processes selected from (1) application of an aqueous solution ⇒ drying ⇒ heat treatment, (2) plating method (wet type), and (3) vapor deposition method (dry type) can be carried out.

[0089] A large diameter ingot is forged, rolled, and drawn to produce a wire (hereinafter also referred to as "intermediate wire") made of Cu or a Cu alloy and having a diameter of about 0.7 to 2.0 mm.

[0090] Although electrolytic plating, electroless plating, vapor deposition, and the like can be used as techniques for forming a coating layer on the surface of a Cu core material, electrolytic plating, which allows for stable control of film thickness, is industrially preferred. For example, a coating layer may be formed on the surface of a large-diameter ingot, a coating layer may be formed on the surface of an intermediate wire, or a coating layer may be formed on the surface of the Cu core material after the intermediate wire is drawn and further thinned (e.g., after being drawn to the final diameter of the Cu core material). Among these, since it is easy to adjust the width of the crystal grains on the wire surface in the above condition (1) to the desired range, it is preferable to form the coating layer when the Cu core material has a large diameter of 50 to 500 times the final wire diameter, and it is preferable to form the coating layer on the surface of a large-diameter ingot. This is because forming the coating layer when the Cu core material has a large diameter can increase the degree of processing of the coating layer in subsequent wiredrawing processes, etc., and facilitates finer crystal grains at the final wire diameter.

[0091] As mentioned above, the coating layer may be formed, for example, by providing an alloy layer containing B and C elements in a predetermined ratio on the surface of a Cu core material, and then providing an A layer on top of that (hereinafter also referred to as "two-stage coating"). The coating layer may also be formed, for example, by providing a C layer on the surface of a Cu core material, then providing a B layer on that, and then providing an A layer on top of that (hereinafter also referred to as "three-stage coating"). The coating layer may also be formed, for example, by providing an alloy layer containing A, B, and C elements in a predetermined ratio on the surface of a Cu core material (hereinafter also referred to as "one-stage coating"). Below, an example of a procedure for forming a coating layer by two-stage coating using Au as the A element and Pd and Ni as the B and C elements is shown. That is, the coating layer may be formed, for example, by providing a PdNi alloy layer containing Pd and Ni in a predetermined ratio on the surface of a Cu core material, and then providing an Au layer on top of that. Here, from the viewpoint of forming a coating layer having excellent adhesion to the Cu core material, a PdNi alloy layer containing Pd and Ni in a predetermined ratio may be provided after strike plating of a conductive metal on the surface of the Cu core material. Alternatively, the coating procedure may be appropriately adjusted, for example, by forming a PdNi alloy layer containing Pd and Ni in a predetermined ratio, and then providing a layer containing one or more of Pd and Ni (for example, a Pd layer, a Ni layer, a PdNi alloy layer), and then providing an Au layer thereon.

[0092] The wiredrawing process can be performed using a continuous wiredrawing device that can accommodate multiple diamond-coated dies. Heat treatment can be performed during the wiredrawing process, if necessary. The heat treatment diffuses the constituent elements of the coating layer into each other, forming a region containing a predetermined concentration of three or more elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag near the surface of the wire (e.g., an alloy region containing Au, Pd, and Ni) so as to satisfy the above condition (2). A preferred method for this is to continuously sweep the wire at a constant speed in an electric furnace at a constant furnace temperature to promote alloying, as this allows for reliable control of the alloy composition and thickness.

[0093] Below, an embodiment in which a coating layer is formed on the surface of a Cu core material by electroplating will be exemplified, with a preferred aspect being taken into consideration in terms of adjusting the width of the crystal grains on the wire surface to the desired range under the above condition (1).

[0094] In such a preferred embodiment, the conditions for electrolytic plating, wire drawing, and heat treatment are controlled to adjust the width of the crystal grains on the wire surface to a desired range.

[0095] First, in the electrolytic plating step, the temperature of the plating bath is preferably set to a low temperature of 60°C or less, more preferably 50°C or less, in order to induce nucleation of the constituent elements of the coating layer and suppress the growth of crystal grains. The lower limit of the plating bath temperature is not particularly limited as long as electrolytic plating can be carried out smoothly, and may be, for example, 10°C or more, 15°C or more, or 20°C or more. In the electrolytic plating step, it is also preferable to stir the plating solution and carry out the plating treatment in a liquid flowing state in order to induce nucleation of the constituent elements of the coating layer and suppress the growth of crystal grains. Therefore, in a preferred embodiment, the electrolytic plating step is carried out in a plating bath in a liquid flowing state at 10 to 60°C (more preferably 10 to 50°C).

[0096] Regarding the wiredrawing and heat treatment, it is preferable to perform intermediate heat treatment during the wiredrawing process when the wire diameter is in the range of 5 to 50 times the final wire diameter. This is because performing an appropriate intermediate heat treatment adjusts the processing strain inside the coating layer, making it easier to adjust the crystal grain size at the final wire diameter. Also, performing an appropriate intermediate heat treatment causes the constituent elements of the coating layer to diffuse and alloy with each other, delaying recrystallization and making it easier to refine the crystal grain size.

[0097] In a particularly preferred embodiment, when the Cu core material has a large diameter of 50 to 500 times the final wire diameter, a coating layer is formed on the surface of the Cu core material by electrolytic plating in a plating bath in a flowing state at 10 to 60° C. Furthermore, the wire is preferably drawn, and intermediate heat treatment is preferably performed during the drawing process when the diameter is in the range of 5 to 50 times the final wire diameter.

[0098] The wire of the present invention can provide good shape stability of the second bonded portion. The wire of the present invention can also provide good bonding reliability of the second bonded portion even in a harsh high-temperature environment. Therefore, the bonding wire of the present invention can be suitably used as a bonding wire for various bonding wires, including bonding wires for automotive devices and power devices.

[0099] [Method for Manufacturing a Semiconductor Device] A semiconductor device can be manufactured by using the bonding wire for a semiconductor device of the present invention to connect electrodes on a semiconductor chip to external electrodes such as electrodes on a lead frame or circuit board.

[0100] In one embodiment, the semiconductor device of the present invention includes a circuit board, a semiconductor chip, and a bonding wire for electrically connecting the circuit board and the semiconductor chip, and the bonding wire is the wire of the present invention.

[0101] 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, a semiconductor device may be configured including a lead frame and a semiconductor chip mounted on the lead frame, as in the semiconductor device described in JP 2020-150116 A.

[0102] Examples of the semiconductor device 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.).

[0103] The present invention will be specifically described below with reference to examples, although the present invention is not limited to the examples shown below.

[0104] (Sample) First, a method for preparing the sample will be described. The Cu used as the raw material for the Cu core material had a purity of 99.99% by mass or more (4N), with the remainder consisting of inevitable impurities. When the first additional element, second additional element, and third additional element were added, the purity of these elements was 99% by mass or more, with the remainder consisting of inevitable impurities, or a master alloy in which these additional elements were blended in high concentrations with Cu was used.

[0105] The core Cu alloy was first prepared by loading raw materials into a graphite crucible and heating them in a high-frequency furnace. ２The materials were melted by heating to 1090-1500°C in an inert atmosphere such as nitrogen gas or Ar gas, and then continuously cast into ingots with diameters of 4-7 mm. Next, a coating layer was formed on the resulting ingot by electrolytic plating. The coating layer was formed by pickling with hydrochloric acid or sulfuric acid to remove the oxide film on the ingot surface, followed by two-stage coating (e.g., Examples 1, 2, 4, 5, etc.), three-stage coating (e.g., Examples 3, 6, 10, etc.), one-stage coating (e.g., Examples 7, 12, 13, etc.), etc. More specifically, Example No. 10 ... In Examples 1, 2, 4, 5, etc., a PtNi alloy layer or a PdNi alloy layer was formed to cover the entire surface of the ingot using a PtNi alloy plating solution containing Pt and Ni in a predetermined ratio, or a PdNi alloy plating solution containing Pd and Ni in a predetermined ratio, and an Au layer was then formed on the PtNi alloy layer or PdNi alloy layer using an Au plating solution. In Examples 3, 6, 10, etc. based on three-step coating, a Pd layer, a Pt layer, or a Ni layer (first layer) was formed to cover the entire surface of the ingot using a Pd, Pt, or Ni plating solution, a Pd layer, a Pt layer, or a Ni layer (second layer; different from the first layer) was formed on the first layer using a Pd, Pt, or Ni plating solution, and an Ag layer or an Au layer (third layer) was formed on the second layer using an Ag or Au plating solution. In Example 1, etc. based on one-step coating, In Examples 7, 12, and 13, a PtNiAg alloy layer, an AuPdNi alloy layer, or an AuPtNi alloy layer was formed so as to cover the entire surface of the ingot using a PtNiAg alloy plating solution containing Pt, Ni, and Ag in a predetermined ratio, an AuPdNi alloy plating solution containing Au, Pd, and Ni in a predetermined ratio, or an AuPtNi alloy plating solution containing Au, Pt, and Ni in a predetermined ratio. The formation of each layer by electroplating was carried out at a plating bath temperature of 20 to 40°C, and the plating solution was stirred to keep the liquid flowing.

[0106] Thereafter, the wire was processed to the final wire diameter by drawing, wire drawing, etc. In order to control the width of the crystal grains near the wire surface, intermediate heat treatment was performed once or twice during processing as necessary. The conditions for the intermediate heat treatment were a heat treatment temperature of 200 to 600°C and a heat treatment time of 1 to 6 seconds for a wire diameter of 0.1 to 0.3 mm. When intermediate heat treatment was performed, the wire was continuously swept and N ２ After processing to the final wire diameter, the wire was continuously swept and N ２ Alternatively, the thermal refining treatment was performed while flowing Ar gas. The thermal refining treatment temperature was 200 to 600°C, the wire feed speed was 20 to 200 m / min, and the thermal refining time was 0.2 to 0.8 seconds. To narrow the grain width, the thermal refining temperature was lowered or the wire feed speed was set to a higher value, and to widen the grain width, the thermal refining temperature was increased or the wire feed speed was set to a lower value.

[0107] (Test and Evaluation Methods) Test and evaluation methods will be described below.

[0108] [Coating layer thickness analysis by Auger electron spectroscopy (AES)] Depth analysis by AES was used to analyze the thickness of the coating layer. Depth analysis by AES involves alternately performing composition analysis and sputtering to analyze changes in composition in the depth direction, and can obtain concentration changes of each element from the wire surface to the depth (center) (so-called depth direction concentration profile).

[0109] First, the bonding wire to be measured was fixed in a straight line on a sample holder. Then, 1) composition analysis of the wire surface was performed by AES, and then 2) sputtering with Ar and 3) composition analysis of the surface after sputtering were repeated to obtain a concentration profile in the depth direction. ＋The analysis was performed using ions and an acceleration voltage of 2 kV. Furthermore, in the surface composition analysis of 1) and 3), the dimensions of the measurement surface and the conditions for the AES composition analysis were as follows. That is, the wire was positioned so that the center of the width of the measurement surface in the direction perpendicular to the wire axis was the center of the width of the measurement surface, and the measurement surface was determined so that the width of the measurement surface was 5% to 15% of the wire diameter. The length of the measurement surface was 5 times the width of the measurement surface. Then, using an AES device (PHI-700 manufactured by ULVAC-PHI), composition analysis of the wire surface was performed at an acceleration voltage of 10 kV, and the concentrations (atomic %) of each element, Pd, Pt, Au, Ni, Ag, and Cu, were determined. When determining the concentration of each element, gas components such as carbon (C), sulfur (S), oxygen (O), and nitrogen (N), as well as non-metallic elements, were not taken into consideration.

[0110] When obtaining a concentration profile in the depth direction by AES depth analysis, the peaks used to detect each element, Pd, Pt, Au, Ni, Ag, and Cu, were as follows. Specifically, focusing on the energy values ​​of the negative peaks (minimum values) of each element in the Auger electron spectrum, the peaks of Au (2022 eV), Pd (333 eV), Pt (1969 eV), Ag (359 eV), Ni (849 eV), and Cu (922 eV) were used. Furthermore, the analysis software (PHI MultiPak) provided with the AES instrument was used to analyze the Auger electron spectrum and calculate the concentration. To improve the accuracy of the analysis, LLS processing was performed as necessary. Specifically, for target samples (wires) having portions containing both Cu and Ni, LLS processing was performed to separate the Cu and Ni peaks. That is, when analyzing the target sample, Auger electron spectra were obtained using pure Cu and pure Ni, and LLS processing was performed using these spectra as data for the reference sample of the element. Furthermore, when the target sample contained a portion containing Cu but not Ni (Cu-containing portion) and a portion containing Ni but not Cu (Ni-containing portion), the Auger electron spectra of the Cu-containing portion and the Ni-containing portion were used as data for the reference sample of the element, and LLS processing was performed in the concentration analysis of the Ni element and the Cu element. Furthermore, LLS processing was performed on target samples containing Au or Pt to reduce background noise. In this case, LLS processing was performed using the waveform of the Auger electron spectrum near the above-mentioned peak energy value of Au or Pt as a reference.

[0111] The concentration profile in the depth direction was obtained for three measurement surfaces spaced apart from each other by 1 mm or more in the longitudinal direction of the wire.

[0112] - Region d from the surface to a depth of 10 nm ０－１０ Average concentration of each element in the depth direction - When obtaining a concentration profile in the depth direction by AES, the measurement point in the depth direction is ０－１０The depth measured by AES analysis is calculated as the product of the sputtering rate and time. Generally, the sputtering rate is determined by the standard sample SiO ２ The depth analyzed by AES is measured using SiO ２ In other words, the unit of thickness of the coating layer is SiO ２ The converted value was used. ２ The measurement points were set to have equal intervals based on the converted value. In the obtained concentration profile in the depth direction, the Pd concentration C Ｐｄ (atomic %) in the region from the wire surface to a depth of 10 nm (i.e., region d ０－１０ ) C for all measurement points Ｐｄ By arithmetically averaging the values ​​of ０－１０ The average concentration of Pd in ​​the region d was calculated. The average concentrations of Pt, Au, Ni, and Ag were also calculated in the region d ０－１０ The concentration of each element was calculated by arithmetically averaging the concentration values ​​of all the measurement points in region d. ０－１０ The average concentration of each element in the test piece was determined by the arithmetic mean value of the values ​​obtained for the three measurement points.

[0113] -Coating layer thickness- In the obtained concentration profile in the depth direction, the concentration profile was confirmed from the wire surface toward the wire center, and the distance Z1 from the wire surface position Z0 to the depth position where the concentration of the core material, Cu, first reached 50 atomic % was determined as the measured coating layer thickness. The thickness of the coating layer was calculated by taking the arithmetic average of the values ​​obtained for three measurement surfaces. In addition, it was confirmed that the total number of measurement points for the coating layer for the wire of the example was 50 to 100.

[0114] In addition, for the wire of the example, when the thickness of the coating layer is dt, it was confirmed that the total concentration of elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag at a depth of dt from the wire surface, i.e., at the boundary between the coating layer and the Cu core, was 45 atomic % or more in the concentration profile in the depth direction of the wire.Furthermore, for the wire of the example, when the thickness of the coating layer is dt, it was confirmed that the total concentration of elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag at a depth of 0.5 dt from the wire surface was 90 atomic % or more in the concentration profile in the depth direction of the wire.

[0115] [Crystalline Analysis of Wire Surface by Electron Backscattered Diffraction (EBSD)] Crystalline analysis of the wire surface was performed by EBSD using the wire surface as the measurement surface.

[0116] -Width of crystal grains on the wire surface- First, the bonding wire to be measured was fixed linearly to a sample holder. Next, the wire was positioned so that the center of the wire width in the direction perpendicular to the wire axis (circumferential direction of the wire) was the center of the width of the measurement surface, and the measurement surface was determined so that the width of the measurement surface was 7 μm and the length of the measurement surface was 15 μm. Then, using an EBSD measurement device (SU-70 manufactured by Hitachi High-Technologies Corporation), measurements were made at a measurement magnification of 15,000 times and a measurement point interval of 0.03 μm. During the measurement, the acceleration voltage was optimized in the range of 15 to 30 kV depending on the surface condition of the sample. Next, analysis software (OIM analysis manufactured by TSL Solutions) attached to the EBSD measurement device was used to analyze the data, excluding measurement points with a CI (Confidence Index) value of less than 0.1. Boundaries where the misorientation between adjacent measurement points was 5 degrees or more were considered to be grain boundaries, and the minimum number of pixels or number of pixels to be recognized as a grain (corresponding to the Minimum size in the Grain Size item in the settings of the attached software) was set to a value of 2 to 5, and the average size of the grains in the circumferential direction of the wire, i.e., the width (nm) of the grains, was determined. For the analysis, crystal data of the element with the highest average concentration in the coating layer was used.

[0117] The calculation of the width of the crystal grains on the wire surface using analysis software is performed by (i) drawing a line in the width direction of the measurement surface (circumferential direction of the wire), determining the size of each crystal grain in the circumferential direction of the wire from the spacing of the crystal grain boundaries on that line, and (ii) arithmetically averaging the size of each crystal grain in the circumferential direction of the wire to calculate the average size of the crystal grains in the circumferential direction of the wire. This was performed for multiple lines (N = 20 to 50) spaced apart in the longitudinal direction of the wire, and the average value was used. Note that the analysis software automatically sets the lines at equal intervals when the number of lines N is set.

[0118] -Crystal orientation of wire surface- In the same manner as above, the bonding wire to be measured was fixed to the sample holder, the measurement surface was determined, and the crystal orientation of the measurement surface was observed to determine the proportion of the <111> crystal orientations in the longitudinal direction of the wire, which had an angular difference of 15 degrees or less with respect to the longitudinal direction of the wire. Note that the crystal orientation measurement by the EBSD method was performed on three measurement surfaces spaced 1 mm or more apart from each other in the longitudinal direction of the wire, and the average value was used.

[0119] [Measurement of element content] The contents of the first additional element, the second additional element, and the third additional element in the wire were detected as the concentration of the elements contained in the entire wire by analyzing a solution obtained by dissolving the bonding wire in a strong acid using an ICP optical emission spectrometer and an ICP mass spectrometer. As the analyzer, an ICP-OES ("PS3520UVDDII" manufactured by Hitachi High-Tech Science Corporation) or an ICP-MS ("Agilent 7700x ICP-MS" manufactured by Agilent Technologies, Inc.) was used.

[0120] [Shape Stability of Second Joint] A sample for evaluating the shape stability of the second joint was prepared by wedge bonding to the lead portion of the lead frame using a commercially available wire bonder. The lead frame was an Fe-42 atomic % Ni alloy lead frame plated with 1 to 3 μm of Ag, and the stage temperature was 200° C., N ２ +5% H ２ Bonding was carried out under a gas flow of 0.5 L / min.

[0121] - Bonding shape (symmetry of fishtail shape) - Regarding the bonding shape of the second bonded portion, 2000 second bonded portions randomly selected from the samples for evaluating the shape stability of the second bonded portion were observed, and the symmetry of the fishtail shape was evaluated. In detail, portions where the fishtail shape was asymmetric with respect to the wire axis or where curling occurred were counted as defective portions, and the bonded shape was evaluated according to the following criteria.

[0122] Evaluation criteria: ◎◎: 1 or less defective spots ◎: 2 to 4 defective spots ○: 5 to 9 defective spots ×: 10 or more defective spots

[0123] - Bonding width (stability of deformation width of fishtail shape) - Regarding the bonding width of the second bond, 2000 second bonded portions randomly selected from the sample for evaluating the shape stability of the second bonded portion were observed, and the standard deviation of the deformation width of the fishtail shape in the direction perpendicular to the axis of the wire was determined and evaluated according to the following criteria.

[0124] Evaluation criteria: ⊚: 3σ is less than 5 μm ◯: 3σ is 5 μm or more and less than 15 μm ×: 3σ is 15 μm or more

[0125] [Bonding Reliability of Second Bonding Portion] The bonding reliability of the second bonding portion was evaluated by a high temperature storage life test (HTSL).

[0126] A sample was wedge-bonded to the lead portion of the lead frame using a commercially available wire bonder, and sealed with a commercially available thermosetting epoxy resin to prepare a sample for testing the bonding reliability of the second bonding portion. The lead frame was an Fe-42 atomic % Ni alloy lead frame plated with 1 to 3 μm of Ag, and the stage temperature was 200°C, N ２ +5% H ２Bonding was performed under a gas flow of 0.5 L / min. The prepared samples for the bonding reliability test of the second bond were exposed to a high-temperature environment at a temperature of 180°C using a high-temperature incubator. The bonding life of the second bond was determined by conducting a pull test on the wedge bond every 500 hours, and the time until the pull strength value became half of the initial pull strength. The pull strength value was the arithmetic average of measurements taken at 50 randomly selected points on the wedge bond. The pull test after the high-temperature storage test was performed after removing the resin by acid treatment to expose the wedge bond. Evaluation was then performed according to the following criteria.

[0127] Evaluation criteria: ⊚: Bonding life of 2000 hours or more ⊚: Bonding life of 1000 hours or more but less than 2000 hours ◯: Bonding life of 500 hours or more but less than 1000 hours ×: Bonding life of less than 500 hours

[0128] [Bonding Reliability of First Bonded Portion] The bonding reliability of the first bonded portion was evaluated by both a Highly Accelerated Temperature and Humidity Stress Test (HAST) and a High Temperature Storage Life Test (HTSL).

[0129] -HAST- A sample was made for testing the bonding reliability of the first bonding part by bonding a ball to an electrode made by depositing a 1.5 μm thick film of an Al-1.0 mass % Si-0.5 mass % Cu alloy on a silicon substrate on a general metal frame using a commercially available wire bonder, and sealing the sample with a commercially available thermosetting epoxy resin. The ball current was set to 30 to 75 mA, the EFO gap was set to 762 μm, and the tail length was set to 254 μm. ２ +5% H ２The bond was formed while flowing gas at a flow rate of 0.4 to 0.6 L / min, and the diameter was 1.5 to 1.9 times the wire diameter. The prepared samples for evaluating the bonding reliability of the first bond were exposed to a high-temperature, high-humidity environment at a temperature of 130°C and a relative humidity of 85% using an unsaturated pressure cooker tester, and a bias of 7 V was applied. The bonding life of the first bond was determined by conducting a shear test on the ball bond every 48 hours, and the time until the shear strength value reached half of the initial shear strength. The shear strength value was the arithmetic mean value of measurements taken at 50 randomly selected locations on the ball bond. The shear test was performed after removing the resin by acid treatment to expose the ball bond. Evaluation was then performed according to the following criteria.

[0130] Evaluation criteria: ◎: Bonding life is 384 hours or more ○: Bonding life is 240 hours or more but less than 384 hours ×: Bonding life is less than 240 hours

[0131] -HTSL- Samples for evaluating the bonding reliability of the first bonded portion, prepared using the same procedure as above, were exposed to an environment at a temperature of 175°C using a high-temperature incubator. The bonding life of the first bonded portion was determined by conducting a shear test on the ball bonded portion every 500 hours, and defining it as the time until the shear strength value became half of the initial shear strength. The shear strength value was the arithmetic mean value of measurements taken at 50 randomly selected points on the ball bonded portion. The shear test after the high-temperature storage test was conducted after removing the resin with an acid treatment to expose the ball bonded portion. Evaluation was then performed according to the following criteria.

[0132] Evaluation criteria: ◎: Bonding life is 2000 hours or more ○: Bonding life is 1000 hours or more but less than 2000 hours ×: Bonding life is less than 1000 hours

[0133] [Loop Shape] The loop shape stability (reproducibility of the loop profile) was evaluated by connecting 30 trapezoidal loops so that the loop length was 2 mm, the height difference at the joint was 250 μm, and the loop height was 200 μm, and the standard deviation of the maximum loop height was used to measure the height using an optical microscope and evaluate it according to the following criteria.

[0134] Evaluation criteria: ⊚: 3σ is less than 20 μm ◯: 3σ is 20 μm or more and less than 30 μm ×: 3σ is 30 μm or more

[0135] [Bonded shape] The bonded shape of the first bonded portion (the crushed shape of the ball) was evaluated by forming a ball using a commercially available wire bonder under the conditions described in the [Bonding reliability of the first bonded portion] section above, crimping the ball to an electrode formed by depositing a 1.5 μm thick film of an Al-1.0 mass % Si-0.5 mass % Cu alloy on a Si substrate, and observing the result from directly above with an optical microscope (number of evaluations N = 100). The crushed shape of the ball was judged as good if it was close to a perfect circle, and as bad if it was oval or petal-like. Evaluation was then performed according to the following criteria.

[0136] Evaluation criteria: ◎: No defects ○: 1 to 3 defects (no practical problems) △: 4 to 5 defects (no practical problems) ×: 6 or more defects

[0137] [Chip Damage] Evaluation of chip damage was performed by forming a ball using a commercially available wire bonder under the conditions described in the above [Bonding Reliability of First Bond] section, crimping the ball to an electrode formed by depositing a 1.5 μm thick Al-1.0 mass % Si-0.5 mass % Cu alloy film on a Si substrate, then dissolving the wire and electrode in a chemical solution to expose the Si substrate, and observing the Si substrate directly below the bond with an optical microscope (number of evaluations N = 50). Evaluation was then performed according to the following criteria.

[0138] Evaluation criteria: ◎: No cracks or bonding traces ○: No cracks, but bonding traces are found in some places (3 places or less) ×: Other than that

[0139] The evaluation results of the Examples and Comparative Examples are shown in Tables 1 and 2.

[0140]

[0141]

[0142] All of the wires of Example Nos. 1 to 36 were provided with a coating layer that satisfied the specific conditions (1) and (2) of the present invention, and it was confirmed that the fishtail shape had good symmetry and small variation in deformation width in the direction perpendicular to the wire axis, resulting in good shape stability of the second joint. Furthermore, it was confirmed that good bonding reliability of the second joint was achieved even in a harsh high-temperature environment of 180°C. Furthermore, it was confirmed that the wires of Example Nos. 1 to 13 and 15 to 36, in which the coating layer had a thickness of 40 nm to 200 nm, and particularly the wires in the preferred range of 50 nm to 150 nm, were likely to exhibit particularly good bonding reliability of the second joint in a harsh high-temperature environment. Furthermore, when the crystal orientation of the wire surface was measured by EBSD, Example No. 15 to 36, in which the proportion of the <111> crystal orientation with an angle difference of 15 degrees or less with respect to the wire longitudinal direction was 30% to 95%. It was confirmed that all of the wires of Examples 1 to 3, 5 to 27, and 29 to 36 produced good loop shapes. Among them, it was confirmed that the wires of Example Nos. 2, 3, 5 to 10, 12, 16 to 21, 23 to 27, 30 to 33, and 36, which had a <111> crystal orientation ratio in the preferred range of 40% to 85%, were particularly likely to produce good loop shapes. In addition, it was confirmed that the wires of Example Nos. 15 to 19, 25, 26, 32, 33, 35, and 36, which contained a total of 1 ppm by mass or more of the first additional element, produced particularly good crimped shapes at the first bonded portion. It was confirmed that the wires of Example Nos. 20 to 26 and 34 to 36, which contained a total of 1 ppm by mass or more of the second additional element, produced particularly good bonding reliability at the first bonded portion in high-temperature, high-humidity environments. It was confirmed that the wires of Example Nos. 20 to 26 and 34 to 36, which contained a total of 0.011 mass% or more of the third additional element, produced particularly good bonding reliability at the first bonded portion in high-temperature, high-humidity environments. It was confirmed that wires Nos. 3, 7, 11, and 27 to 36 provided particularly good bonding reliability at the first bonded portion in a high-temperature environment. On the other hand, it was confirmed that wires Nos. 1 to 12 had coating layers that did not satisfy at least one of the specific conditions (1) and (2) of the present invention, and thus had poor shape stability at the second bonded portion.

[0143] 1 Bonding wire (wire) 2 Measurement surface 10 Second bonded portion 10w Deformation width of second bonded portion (deformation width in a direction perpendicular to the wire axis) X Wire axis (center of wire width) W Wire width (wire diameter) w ａ Measurement surface width l ａ Length of the measuring surface

Claims

1. A bonding wire for a semiconductor device, comprising a core material made of Cu or a Cu alloy and a coating layer containing a conductive metal other than Cu formed on a surface of the core material, The average size of crystal grains in the circumferential direction of the wire, as obtained by analyzing the surface of the wire by an electron backscattered diffraction (EBSD) method, is 35 nm or more and 140 nm or less; In the concentration profile in the depth direction of the wire measured by Auger electron spectroscopy (AES), the region from the surface to a depth of 10 nm (hereinafter referred to as “region d ０－１０ ") contains three or more elements selected from the group consisting of Pd, Pt, Au, Ni, and Ag, and satisfies the concentration conditions of (i) and (ii) below. (i) Area d ０－１０ For at least three elements among the three or more elements contained in the region d ０－１０ The average concentration of each element in the (ii) Area d ０－１０ For all elements among the three or more elements included in the region d ０－１０ The average concentration of each element in the

2. The bonding wire according to claim 1 , wherein the coating layer has a thickness of 40 nm or more and 200 nm or less.

3. 2. The bonding wire according to claim 1, wherein, in a result of measuring the crystal orientation of the surface of the wire by an EBSD method, the proportion of <111> crystal orientations having an angle difference of 15 degrees or less with respect to the longitudinal direction of the wire is 30% or more and 95% or less.

4. The bonding wire according to claim 1, wherein the concentration profile in the depth direction of the wire is obtained by measuring by AES under the following <conditions> while digging in the depth direction from the surface of the wire by Ar sputtering. <Conditions> The center of the wire width is positioned to be the center of the width of the measurement surface, and the width of the measurement surface is 5% to 15% of the wire diameter, and the length of the measurement surface is 5 times the width of the measurement surface.

5. The bonding wire according to claim 1, further comprising one or more elements (hereinafter referred to as “first additive elements”) selected from the group consisting of B, P, In, and Mg, and a total concentration of the first additive element relative to the entire wire is 1 ppm by mass or more and 100 ppm by mass or less.

6. The bonding wire according to claim 1, further comprising one or more elements (hereinafter referred to as “second additive elements”) selected from the group consisting of Se, Te, As, and Sb, and a total concentration of the second additive elements relative to the entire wire is 1 ppm by mass or more and 100 ppm by mass or less.

7. The bonding wire according to any one of claims 1 to 6, comprising one or more elements (hereinafter referred to as "third additive elements") selected from the group consisting of Ga, Ge, and Ag, and a total concentration of the third additive element with respect to the entire wire is 0.011 mass% or more and 1.5 mass% or less.

8. A semiconductor device comprising the bonding wire according to claim 1.