CN107962313B

CN107962313B - Bonding wire for semiconductor device

Info

Publication number: CN107962313B
Application number: CN201711344468.2A
Authority: CN
Inventors: 山田隆; 小田大造; 榛原照男; 大石良; 斋藤和之; 宇野智裕
Original assignee: Nippon Steel and Sumikin Chemical Co Ltd; Nippon Steel Chemical and Materials Co Ltd
Current assignee: Nippon Steel & Sumitomo New Materials Co ltd; Nippon Steel Chemical and Materials Co Ltd
Priority date: 2015-06-15
Filing date: 2016-05-19
Publication date: 2024-03-08
Anticipated expiration: 2036-05-19
Also published as: CN107962313A; JP5985127B1; JPWO2016203899A1

Abstract

A bonding wire for a semiconductor device, which comprises a Cu alloy core material and a Pd coating layer formed on the surface thereof, and which can improve the bonding reliability of a ball bonding portion at a high temperature and has a resistance ratio (= maximum resistance/0.2% resistance) of 1.1 to 1.6. In the result of measuring the crystal orientation of the core material section in the direction perpendicular to the axis of the bonding wire, the average crystal grain diameter of the core material section in the direction perpendicular to the axis of the bonding wire is set to 0.9-1.5 [ mu ] m, and the resistance ratio is set to 1.6 or less by setting the orientation ratio of < 100 > to 30% or more, among the crystal orientations in the longitudinal direction of the wire, the crystal orientation of 15 degrees or less with respect to the longitudinal direction of the wire.

Description

Bonding wire for semiconductor device

The application is that the application date is 2016, 5, 19, the application number is 201680002657.9, and the invention is named: the chinese patent application for "bonding wires for semiconductor devices" is filed separately.

Technical Field

The present invention relates to a bonding wire for a semiconductor device used for connecting an electrode on a semiconductor element to a wiring of a circuit wiring board such as an external lead.

Background

Conventionally, as a bonding wire for a semiconductor device (hereinafter referred to as a bonding wire) for bonding an electrode on a semiconductor element to an external lead, a thin wire having a wire diameter of about 15 to 50 μm has been mainly used. The bonding method of the bonding wire is generally a thermocompression bonding method using ultrasonic waves in combination, and a general bonding apparatus, a capillary tool for connecting the bonding wire to the inside thereof, or the like can be used. The bonding process of the bonding wire is completed by the following procedures: the wire tip is heated and melted by arc heat input, and after forming a Ball (Free Air Ball) by surface tension, the Ball is press-bonded to an electrode of a semiconductor element heated in a range of 150 to 300 ℃ (hereinafter referred to as "Ball bonding"), and then a loop (loop) is formed, and after that, the wire is press-bonded to an electrode on the external lead side (hereinafter referred to as "wedge bonding"). As an electrode on the semiconductor element to be bonded by the bonding wire, an electrode structure in which an alloy film mainly composed of Al is formed on a Si substrate may be used, and an electrode structure in which an Ag-plated layer and/or a Pd-plated layer is applied to an electrode on the external lead side may be used.

Hitherto, although Au has been the main material of bonding wires, work is being advanced to replace Cu mainly for LSI applications. On the other hand, in recent years, in view of the popularization of electric vehicles and hybrid vehicles, there is an increasing demand for the replacement of Au with Cu in vehicle-mounted device applications.

As Cu bonding wires, cu bonding wires using high purity Cu (purity: 99.99 mass% or more) have been proposed (for example, patent document 1). Cu has a disadvantage of being easily oxidized as compared with Au, and has a problem of poor bonding reliability, sphere formation, wedge bonding, and the like. As a method for preventing surface oxidation of Cu bonding wires, a structure in which the surface of a Cu core material is coated with a metal such as Au, ag, pt, pd, ni, co, cr, ti has been proposed (patent document 2). Further, a structure has been proposed in which a surface of a Cu core material is coated with Pd, and a surface of a Pd coating layer is coated with Au, ag, cu, or an alloy thereof (patent document 3).

Prior art literature

Patent literature

Patent document 1 Japanese patent laid-open No. 61-48543

Patent document 2 Japanese patent application laid-open No. 2005-167020

Patent document 3 Japanese patent application laid-open No. 2012-36490

Disclosure of Invention

In-vehicle devices, bonding reliability in a severe high-temperature and high-humidity environment is required as compared with general electronic devices. In particular, the bonding life of the ball bonding portion for bonding the ball portion of the wire to the electrode is the greatest.

As a representative evaluation method for evaluating the bonding reliability of the ball bonding portion under a high-temperature and high-humidity environment, HAST (Highly Accelerated Temperature and Humidity Stress Test) (high-temperature and high-humidity environment exposure test) is known. In the case of evaluating the joining reliability of the ball joint by HAST, the joining life of the ball joint is evaluated by exposing the ball joint for evaluation to a high-temperature and high-humidity environment having a temperature of 130 ℃ and a relative humidity of 85%, and measuring the change with time of the resistance value of the joint or the change with time of the shear strength of the ball joint.

Further, as a method for evaluating the bonding reliability of the ball bonding portion in a high temperature environment of 170 ℃ or higher, HTS (High Temperature Storage Test) (high temperature placement test) can be used. In the case of evaluating the bonding reliability of the ball bonding portion by using HTS, the bonding life of the ball bonding portion is evaluated by measuring the change with time of the resistance value of the ball bonding portion or the change with time of the shear strength of the ball bonding portion with respect to the sample for evaluation exposed to the high temperature environment.

As a result of the study by the present inventors, it was found that, when the bonding wire includes an element imparting connection reliability in a high-temperature environment, such as Ni, zn, rh, in, ir, pt, the bonding reliability of the ball bonding portion in a high-temperature environment of 130 ℃.

Here, the resistance ratio is defined by the following expression (1).

Resistance ratio = maximum resistance/0.2% resistance (1)

In wedge bonding, the bonding wire is strongly deformed. When the wire is work hardened during deformation, the wire after bonding becomes hard, and as a result, the bonding strength of the wedge bonding is lowered. In order to maintain the wedge joint strength, the endurance ratio defined by the above formula (1) is preferably 1.6 or less. However, as a result of the above elements being contained in the wire in order to improve the joining reliability of the ball joint under a high temperature environment, the endurance ratio increases to exceed 1.6. Thus, the engagement strength of the wedge engagement is reduced.

The invention provides a bonding wire for a semiconductor device, which comprises a Cu alloy core material and a Pd coating layer formed on the surface of the Cu alloy core material, can improve the bonding reliability of a ball bonding part at high temperature, and can use the endurance ratio defined by formula (1) to be 1.1-1.6.

That is, the gist of the present invention is as follows.

[1] A bonding wire for a semiconductor device, characterized by comprising a Cu alloy core material and a Pd coating layer formed on the surface of the Cu alloy core material,

the bonding wire contains elements that impart connection reliability in a high temperature environment,

in the result of measuring the crystal orientation of the core material section in the direction perpendicular to the axis of the bonding wire, the orientation ratio of < 100 > of the crystal orientation in the wire length direction (axis direction) with respect to the angle difference in the wire length direction of 15 degrees or less is 30% or more,

the average crystal grain diameter of the core material section in the direction perpendicular to the axis of the bonding wire is 0.9-1.5 mu m.

[2] The bonding wire for a semiconductor device according to the above [1], wherein the resistance ratio defined by the following formula (1) is 1.1 to 1.6.

Resistance ratio = maximum resistance/0.2% resistance (1)

[3] The bonding wire for a semiconductor device according to [1] or [2], wherein the Pd coating layer has a thickness of 0.015 to 0.150. Mu.m.

[4] The bonding wire for a semiconductor device according to any one of [1] to [3], further comprising an alloy skin layer containing Au and Pd on the Pd coating layer.

[5] The bonding wire for a semiconductor device according to [4], wherein the thickness of the alloy skin layer containing Au and Pd is 0.050 μm or less.

[6] The bonding wire for a semiconductor device according to any one of the above [1] to [5], wherein the bonding wire contains at least 1 element selected from Ni, zn, rh, in, ir, pt, and the concentration of the element is 0.011 to 2 mass% in total with respect to the entire wire.

[7] The bonding wire for a semiconductor device according to any one of [1] to [6], wherein the bonding wire contains 1 or more elements selected from Ga and Ge, and the concentration of the elements is 0.011 to 1.5 mass% in total with respect to the entire wire.

[8] The bonding wire for a semiconductor device according to any one of [1] to [7], wherein the bonding wire contains 1 or more elements selected from As, te, sn, sb, bi, se, and the concentration of the elements is 0.1 to 100 mass ppm, sn.ltoreq.10 mass ppm, sb.ltoreq.10 mass ppm, and Bi.ltoreq.1 mass ppm in total with respect to the entire wire.

[9] The bonding wire for a semiconductor device according to any one of [1] to [8], wherein the bonding wire further comprises at least 1 element selected from B, P, mg, ca, la, and the concentration of each element is 1 to 200 mass ppm with respect to the entire wire.

[10] The bonding wire for a semiconductor device according to any one of [1] to [9], wherein Cu is present on the outermost surface of the bonding wire.

[11] The bonding wire for a semiconductor device according to any one of the above [1] to [10], wherein the Cu alloy core material contains 0.1 to 3.0 mass% in total of a metal element of group 10 of the periodic Table, and the Cu concentration at the wire outermost surface is 1 atomic% or more.

According to the present invention, the joining reliability of the ball joint in a high temperature environment can be improved, and the endurance ratio defined by the formula (1) can be 1.1 to 1.6.

Detailed Description

The bonding wire for a semiconductor device of the present invention comprises a Cu alloy core material and a Pd coating layer formed on the surface of the Cu alloy core material. In the present invention, the bonding wire contains an element that imparts connection reliability in a high-temperature environment, and in the result obtained by measuring the crystal orientation of the core material section in the direction perpendicular to the axis of the bonding wire, the orientation ratio of < 100 > of the crystal orientation in the longitudinal direction of the wire, which is 15 degrees or less with respect to the angular difference in the longitudinal direction of the wire, is 30% or more, and the average crystal grain diameter of the core material section in the direction perpendicular to the axis of the bonding wire is 0.9 to 1.5 μm.

A molding resin (epoxy resin) which is a package (package) of a semiconductor device contains chlorine (Cl) in a molecular skeleton. Cl in the molecular skeleton is hydrolyzed under the HAST evaluation condition, namely, a high-temperature and high-humidity environment with the temperature of 130 ℃ and the relative humidity of 85%, and is dissolved as chloride ions (Cl-). In the case of bonding a Cu bonding wire having no Pd coating layer to an Al electrode, if the Cu/Al bonding interface is exposed to a high temperature, cu and Al are interdiffused to form intermetallic compound Cu ₉ Al ₄ 。Cu ₉ Al ₄ Is susceptible to corrosion by halogen such as Cl, and corrosion proceeds due to Cl eluted from the molding resin, resulting in a decrease in bonding reliability. In the case of Cu lines having a Pd coating,since the interface between the Pd-coated Cu wire and the Al electrode has a structure of Cu/Pd-concentrated layer/Al, cu is higher than that of a Cu wire without Pd coating layer ₉ Al ₄ The formation of intermetallic compounds is suppressed, but the bonding reliability in a high-temperature and high-humidity environment required for the in-vehicle device is insufficient.

In contrast, it is considered that if the element for imparting connection reliability in a high-temperature environment is contained as in the present invention, cu is present in the joint portion ₉ Al ₄ The formation of intermetallic compounds tends to be further suppressed.

From the viewpoint of improving the bonding reliability of the ball bonding portion in a high-temperature environment (in particular, the performance of HTS under a condition of 175 ℃ or higher), the concentration of the element that imparts the bonding reliability in a high-temperature environment is preferably 0.011 mass% or more, more preferably 0.030 mass% or more, and even more preferably, relative to the entire wire. The concentration of the element is 0.050 mass% or more with respect to the whole line, 0.070 mass% or more with respect to the whole line, 0.09 mass% or more with respect to the whole line, 0.10 mass% or more with respect to the whole line, 0.15 mass% or more with respect to the whole line, or 0.20 mass% or more with respect to the whole line. The details of the elements that impart connection reliability in a high-temperature environment will be described later.

As described above, the resistance ratio is defined by the following expression (1).

Resistance ratio = maximum resistance/0.2% resistance (1)

In wedge bonding, the bonding wire is strongly deformed. When the wire is work hardened at the time of deformation, the wire after the joining becomes hard, with the result that the joining strength of the wedge joining is lowered. In order to maintain good wedge bonding strength, the endurance ratio defined by the above formula (1) is preferably 1.6 or less. However, as a result of containing an element imparting connection reliability in a high-temperature environment in an amount sufficient to exert an effect in order to improve the joining reliability of the ball joint in a high-temperature environment, the endurance ratio increases to more than 1.6. As a result of the Cu of the core material containing the above elements, it is considered that an increase in the endurance ratio, that is, an increase in the hardness, occurs. Thus, the engagement strength of the wedge engagement is lowered. On the other hand, as a result of the attempt to lower the resistance ratio within the range of the conventional manufacturing method, the resistance ratio becomes smaller than 1.1, and the wedge joining property is poor.

Therefore, a crystal structure capable of maintaining the endurance ratio of formula (1) in an optimum range of 1.1 to 1.6 even if the bonding wire contains the element that imparts connection reliability under a high temperature environment has been studied. The result shows that: when the endurance ratio of formula (1) is to be kept in the optimum range, it is important to control the crystal structure of the core material in the bonding wire, in particular, both of the following: (i) Among the crystal orientations in the wire length direction among the crystal orientations in the results obtained by measuring the crystal orientations of the core material cross section in the direction perpendicular to the axis of the bonding wire, the crystal orientation ratio of < 100 > with respect to the angle difference in the wire length direction of 15 degrees or less (hereinafter, also simply referred to as "< 100 > orientation ratio"), and (ii) the average crystal grain diameter in the core material cross section in the direction perpendicular to the axis of the bonding wire (hereinafter, simply referred to as "average crystal grain diameter"). Specifically, when a bonding wire is manufactured by a normal manufacturing method, it is not possible to make the < 100 > orientation ratio 30% or more and the average crystal grain size 0.9 μm or more and 1.5 μm or less, and as a result, it is found that the endurance ratio becomes smaller than 1.1 or more than 1.6. On the other hand, by the method of manufacturing by drilling as described later, it was found that the average crystal grain diameter in the core material cross section in the direction perpendicular to the axis of the bonding wire was 0.9 to 1.5 μm, and as a result, the endurance ratio of formula (1) was 1.1 to 1.6, with the orientation ratio of < 100 > including the angle difference of 15 degrees or less with respect to the linear length direction being 30% or more, among the crystal orientations in the linear length direction in the core material cross section in the direction perpendicular to the axis of the bonding wire.

When the < 100 > orientation ratio is 30% or more, the work hardening of the wire with deformation at the time of wedge bonding is small, and therefore the endurance ratio can be made 1.6 or less. However, even in this case, when the average crystal grain size is smaller than 0.9 μm, the 0.2% resistance is high (ductility is lacking), and therefore the resistance ratio becomes smaller than 1.1, and the wedge-bonding property is poor. When the average crystal grain size exceeds 1.5. Mu.m, the < 100 > orientation ratio becomes less than 30%, and the 0.2% endurance is low, so that it is estimated that the endurance ratio exceeds 1.6, and the wedge-bonding property is poor.

Further, in the case where the above condition is satisfied with respect to the crystal structure of the wire, if the content of the element imparting connection reliability in a high-temperature environment in the wire is excessive, the endurance ratio may increase. From the viewpoint of achieving a resistance ratio of 1.6 or less, suppressing the hardening of the bonding wire and suppressing the decrease in wedge bonding property, the total concentration of elements that impart connection reliability in a high-temperature environment is preferably 2.0 mass% or less, 1.8 mass% or less, or 1.6 mass% or less, relative to the entire wire.

When an element that imparts connection reliability in a high-temperature environment is to be contained in the bonding wire, the above-described effects of the present invention can be exhibited regardless of the method of containing the element in the Cu core material or the method of containing the element by coating the Cu core material or the wire surface. Since the amounts of these components added are very small, the effect can be exhibited by a wide variety of methods (variation) of addition, regardless of the method used for addition, if the components are contained.

In the bonding wire of the present invention, the Pd coating layer preferably has a thickness of 0.015 μm or more, more preferably from the viewpoint of obtaining a good FAB shape and further improving bonding reliability of the ball bonding portion in a high-temperature and high-humidity environment required for the in-vehicle device: the Pd coating layer has a thickness of 0.02 μm or more, and more preferably: the Pd coating layer has a thickness of 0.025 μm or more, the Pd coating layer has a thickness of 0.03 μm or more, the Pd coating layer has a thickness of 0.035 μm or more, the Pd coating layer has a thickness of 0.04 μm or more, the Pd coating layer has a thickness of 0.045 μm or more, or the Pd coating layer has a thickness of 0.05 μm or more. On the other hand, since the FAB shape decreases even if the Pd coating layer is too thick, the Pd coating layer preferably has a thickness of 0.150 μm or less, more preferably: the Pd coating layer has a thickness of 0.140 [ mu ] m or less, the Pd coating layer has a thickness of 0.130 [ mu ] m or less, the Pd coating layer has a thickness of 0.120 [ mu ] m or less, the Pd coating layer has a thickness of 0.110 [ mu ] m or less, or the Pd coating layer has a thickness of 0.100 [ mu ] m or less.

The definition of the Cu alloy core material and the Pd coating layer of the bonding wire will be described. The boundary between the Cu alloy core material and the Pd coating layer was determined based on the Pd concentration. The position having a Pd concentration of 50 atomic% is defined as a boundary, the region having a Pd concentration of 50 atomic% or more is defined as a Pd coating layer, and the region having a Pd concentration of less than 50 atomic% is defined as a Cu alloy core material. This is because, if the Pd concentration in the Pd coating layer is 50 atomic% or more, the characteristic improvement effect can be obtained by the structure of the Pd coating layer. The Pd coating layer may include a region of a pure Pd layer, and a region where Pd and Cu have a concentration gradient in the depth direction of the wire. The reason why the region having the concentration gradient can be formed in the Pd coating layer is that atoms of Pd and Cu may be diffused due to heat treatment or the like in the manufacturing process. In the present invention, the concentration gradient means that the degree of concentration change in the depth direction is 10mol% or more per 0.1 μm. Further, the Pd coating layer may contain unavoidable impurities.

In the bonding wire of the present invention, the maximum concentration of Pd in the Pd coating layer is preferably 60 at% or more, more preferably 70 at% or more, 80 at% or more, or 90 at% or more, from the viewpoint of further enjoying the effects of the present invention. In the bonding wire of the present invention, the maximum concentration of Pd in the Pd coating layer is preferably 100 atomic% or less, but in the bonding wire of the present invention, the desired effect can be achieved even when the maximum concentration of Pd in the Pd coating layer is less than 100 atomic%, for example, 99.9 atomic% or less, 99.8 atomic% or less, 99.7 atomic% or less, 99.6 atomic% or less, 99.5 atomic% or less, 99.0 atomic% or less, 98.5 atomic% or less, 98 atomic% or less, 97 atomic% or less, 96 atomic% or 95 atomic% or less.

In the bonding wire of the present invention, the thickness of the region having a Pd concentration of 99.0 atomic% or more in the Pd coating layer may be 40nm or less, for example, 35nm or less, 30nm or less, 25nm or less, 20nm or less, 15nm or less, 10nm or less, or 5nm or less.

The bonding wire of the present invention may further have an alloy skin layer containing Au and Pd on the surface of the Pd coating layer. Thus, the bonding wire of the present invention can further improve bonding reliability and can further improve wedge bondability.

The definition of the alloy skin layer containing Au and Pd of the above bonding wire is explained. The boundary between the alloy skin layer containing Au and Pd and the Pd coating layer was determined based on the Au concentration. The position having an Au concentration of 10 atomic% is defined as a boundary, the region having an Au concentration of 10 atomic% or more is defined as an alloy skin layer containing Au and Pd, and the region having an Au concentration of less than 10 atomic% is defined as a Pd coating layer. In addition, even in the region where the Pd concentration is 50 atomic% or more, if Au is present at 10 atomic% or more, it is determined that the alloy skin layer contains Au and Pd. This is because, if the Au concentration is within the above concentration range, an effect of improving the characteristics can be expected from the structure of the Au skin layer. The alloy skin layer containing Au and Pd is an au—pd alloy as a region containing a region where Au and Pd have a concentration gradient in the depth direction of the wire. The reason why the region having such a concentration gradient can be formed in the alloy skin layer containing Au and Pd is because atoms of Au and Pd are diffused due to heat treatment or the like in the manufacturing process. Furthermore, the alloy skin layer containing Au and Pd may also contain unavoidable impurities and Cu.

In the bonding wire of the present invention, the alloy skin layer containing Au and Pd can react with the Pd coating layer to improve the adhesion (bonding) strength between the alloy skin layer containing Au and Pd, the Pd coating layer, and the Cu alloy core material, and to suppress peeling of the Pd coating layer and the alloy skin layer containing Au and Pd at the time of wedge bonding. Thus, the bonding wire of the present invention can further improve wedge bondability. From the viewpoint of obtaining good wedge bonding properties, the thickness of the alloy skin layer containing Au and Pd is preferably 0.0005 μm or more, more preferably: the thickness of the alloy skin layer containing Au and Pd is 0.001 μm or more, the thickness of the alloy skin layer containing Au and Pd is 0.002 μm or more, or the thickness of the alloy skin layer containing Au and Pd is 0.003 μm or more. From the viewpoint of suppressing core misalignment and obtaining a good FAB shape, the thickness of the alloy skin layer containing Au and Pd is preferably 0.050 μm or less, more preferably: the thickness of the alloy skin layer containing Au and Pd is 0.045 μm or less, the thickness of the alloy skin layer containing Au and Pd is 0.040 μm or less, the thickness of the alloy skin layer containing Au and Pd is 0.035 μm or less, or the thickness of the alloy skin layer containing Au and Pd is 0.030 μm or less. The alloy surface layer containing Au and Pd can be formed by the same method as the method for forming the Pd coating layer.

In the present invention, examples of the element that imparts connection reliability In a high-temperature environment include an element of group 9 of the periodic table (Co, rh, ir), an element of group 10 of the periodic table (Ni, pd, pt), an element of group 11 of the periodic table (Ag, au, etc.), an element of group 12 of the periodic table (Zn, etc.), an element of group 13 of the periodic table (Al, ga, in, etc.), an element of group 14 of the periodic table (Ge, sn, etc.), an element of group 15 of the periodic table (P, as, sb, bi, etc.), an element of group 16 of the periodic table (Se, te, etc.), and the like. These elements may be contained in the bonding wire singly or in combination of two or more.

In the present invention, the bonding wire preferably contains at least 1 element selected from Ni, zn, rh, in, ir, pt as an element for imparting connection reliability in a high-temperature environment. The concentration of these elements is preferably 0.011 to 2 mass% in total with respect to the whole wire.

A silane coupling agent is contained in a molding resin (epoxy resin) as a package of a semiconductor device. The silane coupling agent has an effect of improving adhesion between an organic substance (resin) and an inorganic substance (silicon or metal), and therefore, adhesion between the silane coupling agent and a silicon substrate or metal can be improved. Further, in the case where high adhesion is required for a semiconductor or the like mounted on a vehicle, which requires reliability at a higher temperature, a "sulfur-containing silane coupling agent" may be added. The sulfur contained in the molding resin is released when it is used at 175℃or higher (for example, 175℃to 200 ℃). Further, if sulfur released at a high temperature of 175 ℃ or higher contacts Cu, cu corrosion becomes severe, and sulfide (Cu ₂ S), oxide (CuO). If Cu corrosion occurs in a semiconductor device using a Cu bonding wire, ball bonding is particularly preferredThe joining reliability of the joining portion is lowered.

Therefore, the bonding wire contains at least 1 element selected from Ni, zn, rh, in, ir, pt, and the concentration of the element is set to 0.011 to 2 mass% in total with respect to the entire wire, whereby the bonding reliability in a high-temperature environment (particularly, the performance of HTS under a condition of 175 ℃ or higher) can be improved. From the viewpoint of improving the bonding reliability of the ball bonding portion under a high-temperature environment (in particular, the performance of HTS under a condition of 175 ℃ or higher), the total concentration of the elements is preferably 0.011 mass% or more, more preferably 0.050 mass% or more, and even more preferably, with respect to the entire wire. The concentration of the element is 0.070 mass% or more with respect to the whole line, 0.090 mass% or more with respect to the whole line, 0.10 mass% or more with respect to the whole line, 0.15 mass% or more with respect to the whole line, or 0.20 mass% or more with respect to the whole line. In the following description, at least 1 element selected from Ni, zn, rh, in, ir, pt is also referred to as "element M _A ”。

In the present invention, the bonding wire preferably contains 1 or more elements selected from Ga and Ge as elements for imparting connection reliability in a high-temperature environment, and the concentration of the elements is 0.011 to 1.5 mass% in total with respect to the entire wire. Further, 1 or more elements selected from Ga and Ge may be contained in the line instead of the element M _A Or contains at least 1 element selected from Ga and Ge and M _A . In the following description, 1 or more elements selected from Ga and Ge are also referred to as "element M _B ”。

In the case of FAB formation at the ball joint, ga and Ge in the wire are also diffused into the Pd coating layer. It is considered that Ga and Ge present in the Pd dense layer at the interface between Cu and Al in the ball joint portion further enhance the effect of the Pd dense layer in suppressing mutual diffusion of Cu and Al, and as a result, cu which is liable to corrode in a high-temperature and high-humidity environment is suppressed ₉ Al ₄ Is generated. In addition, ga contained in the wireGe has direct Cu resistance ₉ Al ₄ The possibility of an effect being formed.

Further, a spherical portion was formed using a Pd-coated Cu bond wire containing a predetermined amount of at least 1 element selected from Ga and Ge, and FAB was observed by a scanning electron microscope (SEM: scanning Electron Microscope), and as a result, many precipitates having a diameter (Φ) of about several tens of nm were observed on the surface of FAB. When the precipitate was analyzed by energy dispersive X-ray analysis (EDS: energy Dispersive X-ray Spectroscopy), it was confirmed that Ga and Ge were concentrated. From the above-described situation, although the detailed mechanism is not clear, it is considered that the precipitate observed in the FAB exists at the joint interface between the ball and the electrode, and thus the joint reliability of the ball joint is remarkably improved in a high-temperature and high-humidity environment with a temperature of 130 ℃ and a relative humidity of 85%.

The presence site of Ga or Ge is preferably in the Cu alloy core material, but sufficient effects can be obtained even when the Cu alloy core material is contained in the Pd coating layer and/or an alloy skin layer containing Au and Pd, which will be described later. The method of adding Ga and Ge to Cu alloy core material is easy to accurately control concentration, and improves line productivity and quality stability. Further, since a part of Ga and Ge is also contained in the Pd coating layer and/or the alloy skin layer due to diffusion or the like caused by heat treatment, adhesion at each layer interface is improved, and the line productivity can be further improved.

On the other hand, from the viewpoint of obtaining a good FAB shape and suppressing hardening of the bonding wire to obtain a good wedge bonding property, the total concentration of Ga and Ge is 1.5 mass% or less, preferably 1.4 mass% or less, more preferably 1.3 mass% or less, or 1.2 mass% or less, relative to the entire wire.

In the present invention, the bonding wire preferably contains 1 or more elements selected from As, te, sn, sb, bi, se, and the concentration of the elements is 0.1 to 100 mass ppm, sn 10 mass ppm, sb 10 mass ppm, and Bi 1 mass ppm in total with respect to the entire wire. Furthermore, 1 or more elements selected from As, te, sn, sb, bi, se may be contained in the line instead of the element M _A And/or metaElement M _B Or in the presence of element M _A And/or element M _B Comprises at least 1 element selected from As, te, sn, sb, bi, se. In the following description, 1 or more elements selected from As, te, sn, sb, bi, se are also referred to as "element M _C ”。

When the bonding wire contains at least 1 element selected from As, te, sn, sb, bi, se, the concentration of the element is 0.1 to 100 mass ppm, sn is 10 mass ppm or less, sb is 10 mass ppm or less, and Bi is 1 mass ppm or less, based on the entire wire, bonding reliability of the ball bonding portion in a high-temperature and high-humidity environment required for the in-vehicle device can be further improved. In particular, it is preferable to improve the bonding reliability by improving the bonding life of the ball bonding portion in a high-temperature and high-humidity environment having a temperature of 130 ℃ and a relative humidity of 85%. The total concentration of the elements is preferably 0.1 mass ppm or more, more preferably 0.5 mass ppm or more, still more preferably 1 mass ppm or more, and still more preferably, based on the entire line. The concentration of the element is 1.5 mass ppm or more in total with respect to the entire line, 2 mass ppm or more in total with respect to the entire line, 2.5 mass ppm or more in total with respect to the entire line, or 3 mass ppm or more in total with respect to the entire line. On the other hand, from the viewpoint of obtaining a good FAB shape, the total concentration of the elements is preferably 100 mass ppm or less, more preferably 95 mass ppm or less, 90 mass ppm or less, 85 mass ppm or less, or 80 mass ppm or less, with respect to the entire line. In addition, when the Sn concentration, sb concentration exceeds 10 mass ppm, or Bi concentration exceeds 1 mass ppm, the FAB shape becomes poor, and therefore, setting to 10 mass ppm or less of Sn, 10 mass ppm or less of Sb, and 1 mass ppm or less of Bi is preferable, since the FAB shape can be further improved.

The bonding wire of the present invention preferably further comprises at least 1 element selected from B, P, mg, ca, la, and the concentration of each element is 1 to 200 mass ppm with respect to the whole wire. This can improve the crushed shape of the ball joint required for high-density mounting, that is, can improve the circularity of the ball joint shape. The concentrations of the elements are preferably 1 mass ppm or more, respectively, with respect to the entire line, more preferably: the concentration of the element is 2 mass ppm or more with respect to the entire line, 3 mass ppm or more with respect to the entire line, 4 mass ppm or more with respect to the entire line, or 5 mass ppm or more with respect to the entire line, respectively. From the viewpoint of suppressing chip damage at the time of ball bonding by suppressing hardening of the balls, the concentrations of the elements are preferably 200 mass ppm or less, more preferably 150 mass ppm or less, 120 mass ppm or less, 100 mass ppm or less, 95 mass ppm or less, 90 mass ppm or less, 85 mass ppm or less, or 80 mass ppm or less, respectively, with respect to the entire wire.

As in the present invention, when the Pd-coated Cu bonding wire contains an element that improves connection reliability in a high-temperature environment, cu is present in the bonding portion when Cu is also present on the outermost surface of the bonding wire ₉ Al ₄ The formation of intermetallic compounds tends to be further suppressed. In the case where the Pd-coated Cu bonding wire contains an element that improves connection reliability in a high-temperature environment, when Cu is still present on the outermost surface of the bonding wire, pd concentration on the FAB surface is promoted at the time of FAB formation by interaction between the Cu and the above-mentioned element contained in the bonding wire, and Pd concentration at the ball bonding interface is more remarkably exhibited. It is estimated that the effect of the Pd dense layer to suppress the mutual diffusion of Cu and Al is further enhanced, and Cu which is easily corroded by Cl is present ₉ Al ₄ The amount of produced ball joints is reduced, and the joint reliability of the ball joints in a high-temperature and high-humidity environment is further improved.

When Cu is present on the outermost surface of the Pd coating layer, if the Cu concentration is 30 atomic% or more, the sulfidation resistance of the wire surface is lowered, and the service life of the bonding wire is reduced, which may be not practical. Therefore, in the case where Cu is present at the outermost surface of the Pd coating layer, the concentration of Cu is preferably less than 30 at%.

In addition, when Cu is present on the outermost surface of the Au skin layer, if the Cu concentration is 35 at% or more, the sulfidation resistance of the wire surface is lowered, and the service life of the bonding wire is reduced, which may not be practical. Therefore, in the case where Cu is present at the outermost surface of the Au skin layer, the concentration of Cu is preferably less than 35 at%.

Here, the outermost surface refers to a region measured by an auger electron spectroscopy device on the surface of the bonding wire without sputtering or the like.

In the present invention, the Cu alloy core material preferably contains 0.1 to 3.0 mass% in total of a metal element of group 10 of the periodic table, and the Cu concentration at the line outermost surface is 1 to 10 atomic%. With such a configuration, the wedge bonding property of the Pd-plated lead frame or the lead frame having Au plating on the Pd plating layer can be further improved. Further, by containing a predetermined amount of a metal element of group 10 of the periodic table in the Cu alloy core material, excellent ball bonding properties under high-humidity heating conditions can be achieved with respect to the ball bonding portion between the bonding wire and the electrode.

The metal element of group 10 of the periodic table in the Cu alloy core is preferably 1 or more selected from Ni, pd and Pt. In a preferred embodiment, the Cu alloy core material contains Ni as a metal element of group 10 of the periodic table. For example, the Cu alloy core material may contain Ni alone or may contain one or both of Pd and Pt in combination with Ni as the metal element of group 10 of the periodic table. In another preferred embodiment, the Cu alloy core material contains one or both of Pd and Pt as a metal element of group 10 of the periodic table.

If the concentration of the metal elements of group 10 of the periodic table in the Cu alloy core material is 0.1 mass% or more in total, the interdiffusion of Cu and Al at the joint interface can be sufficiently controlled, and the life of the joint portion is also improved to 380 hours or more in a HAST test which is a severe high-humidity heating evaluation test. As the evaluation of the joint, the resin was unsealed and removed after the HAST test, and then the fracture state of the joint was evaluated by a pull (pull) test. From the viewpoint of sufficiently obtaining the above-described improvement effect of HAST test reliability, the concentration of the metal element of group 10 of the periodic table in the Cu alloy core material is 0.1 mass% or more, preferably 0.2 mass% or more, more preferably: the concentration of the metal elements of group 10 of the periodic table in the Cu alloy core material is 0.3 mass% or more in total, the concentration of the metal elements of group 10 of the periodic table in the Cu alloy core material is 0.4 mass% or more in total, the concentration of the metal elements of group 10 of the periodic table in the Cu alloy core material is 0.5 mass% or more in total, or the concentration of the metal elements of group 10 of the periodic table in the Cu alloy core material is 0.6 mass% or more in total. Further, the concentration of the metal element of group 10 of the periodic table in the Cu alloy core material is 3.0 mass% or less, preferably 2.5 mass% or less, or 2.0 mass% or less in total, from the viewpoints of obtaining a bonding wire excellent in initial bonding strength with an Al electrode in low-temperature bonding, long-term reliability in HAST test, and mass production margin for bonding to substrates such as BGA (Ball Grid Array) and CSP (Chip Size Package), a tape, and the like, and reducing chip damage. When the concentration of the metal elements of group 10 of the periodic table in the Cu alloy core material exceeds 3.0 mass%, ball bonding is required at a low load to avoid chip damage, and the initial bonding strength with the electrode is lowered, and as a result, the HAST test reliability may be deteriorated. In the bonding wire of the present invention, the reliability in the HAST test is further improved by setting the total concentration of the metal elements of group 10 of the periodic table in the Cu alloy core material to the above-described optimum range. For example, a bonding wire having a life exceeding 450 hours until failure occurs in the HAST test can be realized. This is also sometimes longer than 1.5 times or more of the conventional Cu bonding wire, and can be used in severe environments.

Further, as a method for determining the concentration of the above-mentioned element contained in the Cu alloy core material from the bonding wire product, for example, there is mentioned: a method of exposing the section of the bonding wire and analyzing the concentration of the region of the Cu alloy core material; a method of analyzing the concentration of the Cu alloy core material region while cutting the surface of the bonding wire in the depth direction by sputtering or the like. For example, in the case where the Cu alloy core material includes a region having a Pd concentration gradient, a line analysis may be performed on a cross section of the joint, and a concentration analysis may be performed on a region having no Pd concentration gradient (for example, a region having a Pd concentration change in the depth direction of less than 10mol% per 0.1 μm, or an axial center portion of the Cu alloy core material).

In the bonding wire of the present invention, by using a Cu alloy core material containing a predetermined amount of a metal element of group 10 of the periodic table, and setting the Cu concentration at the wire outermost surface to 1 atom% or more, wedge bonding, particularly peeling (peeling), to a Pd-plated lead frame can be significantly improved, good wedge bonding and FAB shape can be realized, oxidation of the wire surface can be suppressed, and deterioration of quality with time can be suppressed. In the bonding wire of the present invention, the Cu concentration at the wire outermost surface is preferably 1.5 at% or more, more preferably 2 at% or more, 2.5 at% or more, or 3 at% or more, from the viewpoint of further improving the wedge bonding property. The upper limit of the Cu concentration at the wire outermost surface is as described above, but in the bonding wire of the present invention including the Cu alloy core material containing a predetermined amount of the metal element of group 10 of the periodic table, the Cu concentration at the wire outermost surface is preferably 10 at% or less, more preferably 9.5 at% or less, or 9 at% or less from the viewpoint of achieving good wedge bonding property and FAB shape, and suppressing oxidation of the wire surface to suppress deterioration of quality with time.

The concentration analysis of the Pd coating layer and the alloy skin layer containing Au and Pd is effective as a method of analyzing while cutting by sputtering or the like from the surface of the bonding wire in the depth direction, or as a method of exposing the wire cross section to perform wire analysis, dot analysis, or the like. As the analyzer used for these concentration analyses, an auger electron spectroscopy analyzer, an energy dispersive X-ray analyzer, an electron beam microscopy analyzer, or the like, which is equipped in a scanning electron microscope or a transmission electron microscope, can be used. As a method for exposing the line section, mechanical polishing, ion etching, or the like can be used. The analysis of trace elements such as Ni, zn, rh, in, ir, pt in the bonding wire can be performed by analyzing a liquid obtained by dissolving the bonding wire with a strong acid by using an ICP emission spectrometry device or an ICP mass spectrometry device, and detecting the liquid as the concentration of the element contained in the whole bonding wire.

(manufacturing method)

Next, a method for manufacturing a bonding wire according to an embodiment of the present invention will be described. The bonding wire is obtained by manufacturing a Cu alloy for a core material, processing the Cu alloy into a thin wire shape, forming a Pd coating layer and an Au layer, and performing a heat treatment. After forming the Pd coating layer and the Au layer, the wire drawing and heat treatment may be performed again. A method for producing the Cu alloy core material, a Pd coating layer, a method for forming an alloy skin layer containing Au and Pd, and a heat treatment method will be described in detail.

The Cu alloy used for the core material is obtained by melting Cu as a raw material together with an added element and solidifying the same. For the melting, an arc heating furnace, a high-frequency heating furnace, a resistance heating furnace, or the like can be used. To prevent mixing of O from the atmosphere ₂ 、H ₂ The isogas is preferably in a vacuum atmosphere or Ar, N ₂ And melting in an inert atmosphere.

Methods for forming a Pd coating layer and an Au layer on the surface of the Cu alloy core material include plating, vapor deposition, and melting. As the plating method, any of an electrolytic plating method and an electroless plating method can be applied. Electrolysis plating, called strike plating or flash plating, has a high plating speed and is excellent in adhesion to a substrate. Solutions used for electroless plating can be classified into a substitution type and a reduction type, and in the case of a thin thickness, substitution type plating alone is sufficient, but in the case of a thick thickness, it is effective to perform reduction type plating stepwise after substitution type plating.

In the vapor deposition method, physical adsorption such as sputtering, ion plating, vacuum vapor deposition, and chemical adsorption such as plasma CVD can be used. The dry method is not required to be washed after the Pd coating layer and the Au layer are formed, and there is no concern about surface contamination during washing.

By performing heat treatment after the formation of the Pd coating layer and the Au layer, pd in the Pd coating layer diffuses into the Au layer, thereby forming an alloy skin layer containing Au and Pd. Instead of forming the Au layer and then performing heat treatment to form the Au-and Pd-containing alloy skin layer, the Au-and Pd-containing alloy skin layer may be covered from the beginning.

Any of a method of forming a Pd coating layer, an alloy skin layer containing Au and Pd, and a method of forming after drawing to a final wire diameter, and a method of drawing several times after forming on a Cu alloy core material of a large diameter until a target wire diameter is effective. In the former case, when a Pd coating layer and an alloy skin layer containing Au and Pd are formed at the final wire diameter, manufacturing, quality control, and the like are easy. In the latter case, the combination of the Pd coating layer, the alloy skin layer containing Au and Pd, and the wire drawing is advantageous in improving the adhesion with the Cu alloy core material. Specific examples of the respective formation methods include: a method of forming a Pd coating layer and an alloy skin layer containing Au and Pd on a Cu alloy core material having a final wire diameter while continuously sweeping the wire through an electrolytic plating solution; alternatively, a method in which a crude Cu alloy core material is immersed in an electrolytic plating bath or an electroless plating bath to form a Pd coating layer and an alloy skin layer containing Au and Pd, and then the wire is drawn to a final wire diameter; etc.

After forming the Pd coating layer and the alloy skin layer containing Au and Pd, heat treatment is sometimes performed. By performing the heat treatment, atoms are diffused between the alloy skin layer containing Au and Pd, the Pd coating layer, and the Cu alloy core material, and the adhesion strength is improved, so that peeling of the alloy skin layer containing Au and Pd and the Pd coating layer during processing can be suppressed, and the productivity is effectively improved. To prevent O from the atmosphere ₂ Preferably in a vacuum atmosphere or Ar, N ₂ And performing heat treatment in an inert atmosphere.

As described above, by adjusting the conditions of the diffusion heat treatment and the annealing heat treatment performed on the bonding wire, cu of the core material is diffused in the Pd coating layer and the alloy skin layer containing Au and Pd by grain boundary diffusion, intra-grain diffusion, or the like, so that Cu can reach the outermost surface of the bonding wire and Cu can be present on the outermost surface. As the heat treatment for making Cu exist on the outermost surface, as described above, a heat treatment for forming an alloy skin layer containing Au and Pd can be used. When the heat treatment for forming the alloy skin layer is performed, cu can be present on the outermost surface or Cu can be absent on the outermost surface by selecting the heat treatment temperature and time. Further, the Cu concentration at the outermost surface can be adjusted to a predetermined range (for example, a range of 1 to 50 atomic%). Cu may be diffused to the outermost surface by a heat treatment other than the formation of the alloy skin layer.

As described above, when an element that imparts connection reliability in a high-temperature environment is contained in a bonding wire, the above-described effects of the present invention can be exhibited regardless of the method of containing the element in a Cu core material or the method of containing the element by coating the Cu core material or the wire surface. The same applies to B, P, mg, ca, la.

The most convenient method for adding the above components is to add the components to the starting material of the Cu alloy core material in advance. For example, after weighing high-purity copper and the component element raw materials as starting materials, they are melted by heating under high vacuum or an inert atmosphere such as nitrogen or argon, thereby producing an ingot to which the component in the target concentration range is added as a starting material containing the component element in the target concentration. Therefore, in a preferred embodiment, the Cu alloy core material of the bonding wire of the present invention contains at least 1 element selected from Ni, zn, rh, in, ir, pt such that the concentration of the above elements is 0.011 to 2 mass% in total with respect to the entire wire. Suitable numerical ranges for the concentration totals are as previously described. In another preferred embodiment, the Cu alloy core material of the bonding wire of the present invention contains 1 or more elements selected from Ga and Ge such that the concentration of the elements is 0.011 to 1.5 mass% in total with respect to the entire wire. Suitable numerical ranges for the concentration totals are as previously described. In another preferred embodiment, the Cu alloy core material of the bonding wire of the present invention contains at least 1 element selected from As, te, sn, sb, bi, se such that the total concentration of the above elements is 0.1 to 100 mass ppm, 10 mass ppm or less of Sn, 10 mass ppm or less of Sb, and 1 mass ppm or less of Bi with respect to the entire wire. Suitable numerical ranges for this concentration are as previously described. In a preferred embodiment, the Cu purity of the Cu alloy core material is 3N or less (preferably 2N or less). The conventional Pd-coated Cu bonding wire tends to use a Cu core material having high purity (4N or more) and avoid using a Cu core material having low purity from the viewpoint of bondability (bondability). The bonding wire of the present invention containing a specific element is particularly suitable for use with a Cu alloy core material having low Cu purity as described above, and achieves bonding reliability of a ball bonding portion in a high-temperature and high-humidity environment required for a vehicle-mounted device. In another preferred embodiment, the Cu alloy core material of the bonding wire of the present invention contains at least 1 element selected from B, P, mg, ca, la such that the concentration of the element is 1 to 200 mass ppm, respectively, with respect to the entire wire. Suitable numerical ranges for this concentration are as previously described. In a preferred embodiment, the Cu alloy core material of the bonding wire of the present invention contains 0.1 to 3.0 mass% in total of a metal element of group 10 of the periodic table. Suitable numerical ranges for this concentration are as previously described.

The above components may be contained by coating the wire surface during the wire manufacturing process. In this case, the coating step may be incorporated at any time point in the line manufacturing process, and the coating may be repeated a plurality of times. Can be combined into a plurality of procedures. The metal may be added to the Cu surface before Pd coating, may be added to the Pd surface after Pd coating, may be added to the Au surface after Au coating, and may be incorporated into each coating step. As a coating method, one can use the coating method from (1) (2) Plating (wet) and (3) vapor deposition (dry).

In the adoption ofIn the case of the method of (2), an aqueous solution of an appropriate concentration is first prepared with a water-soluble compound containing the above-mentioned constituent elements. This allows the above components to be incorporated into the wire material. The coating step may be incorporated at any point in the line manufacturing process, and the coating may be repeated a plurality of times. Can be combined into a plurality of procedures. Can be added to the Cu surface before Pd coatingThe metal oxide may be added to the Pd surface after Pd coating, or may be added to the Au surface after Au coating, or may be incorporated into each coating step.

In the case of using the plating method (wet method), the plating method can be any of an electrolytic plating method and an electroless plating method. In addition to normal electrolytic plating, a plating method called flash plating, which is fast in plating speed and excellent in adhesion to a substrate, can be applied to the electrolytic plating method. Solutions for electroless plating are substitution type and reduction type. In general, substitution plating is applicable when the plating thickness is small, and reduction plating is applicable when the plating thickness is large, but any plating solution concentration and time may be selected and adjusted according to the concentration to be added. Both the electrolytic plating method and the electroless plating method may be incorporated at any time point in the online production process, and may be repeated a plurality of times. Can be combined into a plurality of procedures. The metal may be added to the Cu surface before Pd coating, may be added to the Pd surface after Pd coating, may be added to the Au surface after Au coating, and may be incorporated into each coating step.

Examples of the vapor deposition method (dry method) include sputtering, ion plating, vacuum vapor deposition, and plasma CVD. Since it is dry, pretreatment and post-treatment are not required, and there is no fear of contamination, which is an advantage. Generally, the vapor deposition method has a problem that the addition rate of the aimed element is low, but is one of the methods suitable for the purpose of the present invention because the addition concentration of the above-mentioned component elements is relatively low.

Each vapor deposition method may be incorporated at any time point in the online production process, and may be repeated a plurality of times. Can be combined into a plurality of procedures. The metal may be added to the Cu surface before Pd coating, may be added to the Pd surface after Pd coating, may be added to the Au surface after Au coating, and may be incorporated into each coating step.

The production method for measuring the crystal orientation of the core material cross section in the direction perpendicular to the axis of the bonding wire will be described, wherein the orientation ratio of < 100 > of the crystal orientation in the longitudinal direction of the wire, which is 15 degrees or less with respect to the angular difference in the longitudinal direction of the wire, is 30% or more, and the average crystal grain diameter of the core material cross section in the direction perpendicular to the axis of the bonding wire is 0.9 to 1.5 μm.

When the Cu alloy core material is made to contain an element that imparts connection reliability in a high-temperature environment as a bonding wire, the material strength (hardness) of the wire is increased. Therefore, when the bonding wire of the Cu core wire is wire-drawn, the surface reduction ratio at the time of wire drawing is set to be 5 to 8% lower. Further, since the hardness is still high in the heat treatment after drawing, the heat treatment is performed at 600 ℃ or higher in order to soften the wire to a level that can be used as a bonding wire. Since the heat treatment temperature is high, the < 100 > orientation ratio in the wire longitudinal direction becomes less than 30%, the average crystal grain diameter of the core material cross section becomes more than 1.5 μm, and the endurance ratio becomes more than 1.6. On the other hand, when the heat treatment temperature is lowered to lower the resistance ratio, the average crystal grain diameter of the core material cross section becomes smaller than 0.9 μm, the resistance ratio becomes smaller than 1.1, and the wedge-bonding property is poor.

In contrast, in the present invention, when drawing using a die, the reduction ratio is 10% or more for half or more of the total die number, and the heat treatment temperature in the heat treatment after drawing is set to a temperature of 500 ℃ or less. As a result, in the results obtained by measuring the crystal orientation of the core material cross section in the direction perpendicular to the axis of the bonding wire, the orientation ratio of < 100 > of the crystal orientation in the wire longitudinal direction with respect to the crystal orientation in the wire longitudinal direction, which is 15 degrees or less, can be set to 30% or more, and the average crystal grain diameter of the core material cross section in the direction perpendicular to the axis of the bonding wire can be set to 0.9 to 1.5 μm. By applying the latest drawing technique, and by designing the concentration of the nonionic surfactant contained in the lubricating liquid to be higher than before as the lubricating liquid, the approach angle (approach angle) of the drawing die to be gentler than before as the drawing die shape, and by setting the cooling water temperature of the drawing die to be lower than before, the synergistic effect is achieved, and even though the Cu alloy core material is hardened by containing Ni or other components in total of 0.03 mass% or more, the drawing process with a reduction ratio of 10% or more can be realized.

In measuring the crystal orientation of the line section, an electron back scattering diffraction method (EBSD: electron Backscattered Diffraction) is preferably used. The EBSD method is characterized in that it can observe the crystal orientation of the observation surface and can show the angle difference of the crystal orientation between adjacent measurement points, and can observe the crystal orientation with relatively ease and high accuracy even in a thin line such as a bonding wire. The particle size measurement can be obtained by using analysis software provided in the apparatus for measuring the result obtained by the EBSD method. The crystal grain size (crystal grain size) specified in the present invention is a value obtained by arithmetically averaging the equivalent diameter of crystal grains (diameter of circle corresponding to area of crystal grains; diameter of circle corresponding to area of crystal grains) contained in the measurement region.

The present invention is not limited to the above-described embodiments, and can be appropriately modified within the scope of the gist of the present invention.

Examples

The bonding wire according to the embodiment of the present invention will be specifically described below with reference to examples.

Examples 1 to 59 of the present invention and comparative examples 1 to 16 >

(preparation of sample)

First, a method for producing a sample will be described. Cu, which is a material of the core material, has a purity of 99.99 mass% or more, and the remainder is composed of unavoidable impurities. Au, pd, ni, zn, rh, in, ir, pt a material having a purity of 99 mass% or more and the balance of unavoidable impurities is used. The element (Ni, zn, rh, in, ir, pt) added to the core material is formulated so that the composition of the wire or core material reaches the target composition. The addition of Ni, zn, rh, in, ir, pt may be performed by simple substance, but in the case of an element having a high melting point as a simple substance and/or an addition amount of an extremely small amount, a Cu master alloy containing an addition element may be prepared in advance and blended so as to be a target addition amount. In the present invention examples 27 to 47, 1 or more kinds of Ga, ge, as, te, sn, sb, bi, se, B, P, mg, ca, la were contained.

Cu alloy of core materialThe wire diameter is made by continuous casting so as to be several millimeters. The resulting alloy was drawn to produce a wire having a diameter of 0.3 to 1.4 mm. A commercially available lubricant was used for drawing, and the drawing speed was set to 20 to 150 m/min. In order to remove the oxide film on the wire surface, acid washing treatment with hydrochloric acid or the like is performed, and then a Pd coating layer having a thickness of 1 to 15 μm is formed so as to cover the entire surface of the Cu alloy of the core material. Further, a part of the wires forms an alloy skin layer containing Au and Pd with a thickness of 0.05 to 1.5 μm on the Pd coating layer. For the formation of the Pd coating layer and the alloy skin layer containing Au and Pd, an electrolytic plating method was used. The plating solution used was a commercially available semiconductor plating solution. Then, the drawing process is mainly performed using a drawing die having a reduction of 10 to 21%, and further, the drawing process is performed for 1 to 3 times at a temperature of 200 to 500 ℃ while the drawing process is being performed until the drawing process reaches a diameter of 20. Mu.m. After processing, heat treatment is finally performed so that the elongation at break becomes about 5 to 15%. The heat treatment method is to circulate N while continuously sweeping the wire ₂ Or Ar gas. The wire conveying speed is set to 10-90 m/min, the heat treatment temperature is 350-500 ℃, and the heat treatment time is set to 1-10 seconds.

(evaluation method)

The content of Ni, zn, rh, in, ir, pt, ga, ge, as, te, sn, sb, bi, se, B, P, mg, ca, la in the wire was analyzed as the concentration of the element contained in the entire bonding wire by an ICP emission spectrometry device.

The concentration analysis of the Pd coating layer and the alloy surface layer containing Au and Pd was performed by auger electron spectroscopy while cutting the surface of the bonding wire in the depth direction by sputtering or the like. From the obtained depth-direction concentration profile, the thickness of the Pd coating layer, the maximum concentration of Pd, and the thickness of the alloy skin layer containing Au and Pd were obtained.

The crystal orientation ratio of < 100 > with respect to the line length direction angle difference of 15 degrees or less among the crystal orientations in the line length direction in the core material cross section in the direction perpendicular to the line axis of the bonding line was calculated after observing the crystal orientation of the observation plane (i.e., the core material cross section in the direction perpendicular to the line axis) by the EBSD method. For analysis of the EBSD measurement data, dedicated software (TSL, OIM analysis, etc.) is used. The average crystal grain size in the cross section of the core material in the direction perpendicular to the axis was calculated after observing the crystal orientation of the observation plane by the EBSD method. Dedicated software (TSL solver, OIM analysis, etc.) is used for analyzing EBSD measurement data. The crystal grain size is a value obtained by arithmetically averaging the equivalent diameter of crystal grains (diameter of a circle corresponding to the area of the crystal grains; equivalent circle diameter) contained in the measurement region.

Regarding the 0.2% endurance and the maximum endurance, evaluation was performed by a tensile test with the distance between punctuation set to 100 mm. As a means for the tensile test of the steel sheet, a model 5542 universal material tester manufactured by the company "i" was used. The 0.2% endurance was calculated using dedicated software equipped in the device. The load at break was set as the maximum endurance. The resistance ratio is calculated from the following expression (1).

Resistance ratio = maximum resistance/0.2% resistance (1)

With respect to the evaluation of wedge bondability in the wire bonding portion, 1000 bonds were performed to the wedge bonding portion of the BGA substrate, and the judgment was made based on the occurrence frequency of peeling of the bonding portion. The BGA substrate used was a substrate plated with Ni and Au. In this evaluation, the stage temperature was set to 150 ℃. In the above evaluation, when 11 or more failures occurred, it was judged that the failure was marked as x, and if the failure was 6 to 10, it was judged that the failure was practical but that the failure was slightly problematic, and when the failure was 1 to 5, it was judged that the failure was no problem, and when the failure was not occurred, it was judged that the failure was excellent, and the failure was marked as excellent, which are all described in the column "wedge-bonding property" in tables 1 to 4.

The bonding reliability of the ball bonding portion in a high-temperature and high-humidity environment or a high-temperature environment was determined by preparing a sample for evaluating bonding reliability, performing HTS evaluation, and determining the bonding life of the ball bonding portion. The samples for evaluating the bonding reliability are those obtained in the general casesAn electrode formed by forming an Al-1.0% Si-0.5% Cu alloy film having a thickness of 0.8 μm on a Si substrate on a metal frame, ball bonding by a commercially available wire bonder, and packaging by a commercially available epoxy resin. The ball circulates N at a flow rate of 0.4 to 0.6L/min ₂ +5％H ₂ The gas is formed while having a size in the range of from 33 to 34. Mu.m.

For HTS evaluation, a high temperature thermostat was used, and the produced samples for joint reliability evaluation were exposed to a high temperature environment at 200 ℃. The joint life of the ball joint was set to a time period in which the shear strength of the ball joint was 1/2 of the initial shear strength obtained by conducting the shear test every 500 hours. The shear test after the high temperature and high humidity test is performed after the resin is removed by acid treatment to expose the ball joint.

A shear tester manufactured by DAGE company was used for the HTS evaluation. The shear strength values were averaged over the measured values at 10 points of the randomly selected ball joint. In the above evaluation, if the joining life is less than 500 hours, it is judged that the joining life is not practical and marked as x, if it is 500 hours or more and less than 1000 hours, it is judged that the joining life is practical but needs improvement and marked as Δ, if it is 1000 hours or more and less than 3000 hours, it is judged that the joining life is practically no problem and marked as o, and if it is 3000 hours or more, it is judged that the joining life is particularly excellent and marked as o, and all of them are described in the columns "HTS" in tables 1 to 4.

The ball formability (FAB shape) was evaluated by observing the ball before joining, and judging whether or not there was air bubbles on the ball surface and deformation of the ball which was originally a sphere. If any of the above-described phenomena occurs, the failure is determined. In order to suppress oxidation in the melting step, N was blown at a flow rate of 0.5L/min for the formation of balls ₂ The gas is carried out at the same time. The size of the spheres was set to 34 μm. 50 balls were observed for 1 condition. SEM was used for observation. In the evaluation of the sphericity, if 5 or more failures occur, it is judged that there are problems and marked as x, and if 3 to 4 failures are detected, it is judged that the method is practical but there are problems and marked as delta, and if the number of failures is 1 to 2If the number is "good", the "FAB shape" columns in tables 1 to 4 are all written.

The joining life of the ball joint in a high-temperature and high-humidity environment at 130 ℃ and a relative humidity of 85% can be evaluated by the following HAST evaluation. For HAST evaluation, a sample for evaluating the bonding reliability was exposed to a high-temperature and high-humidity environment having a temperature of 130 ℃ and a relative humidity of 85% using an unsaturated pressure cooker retort tester, and a bias voltage of 5V was applied. The joint life of the ball joint was set to a time period of 1/2 of the initial shear strength obtained by conducting the shear test of the ball joint every 48 hours. The shear test after the high temperature and high humidity test is performed after the resin is removed by acid treatment to expose the ball joint.

As a shear tester for HAST evaluation, a tester manufactured by DAGE company was used. The shear strength values were averaged over the measured values at 10 points of the randomly selected ball joint. In the above evaluation, if the joining life is less than 144 hours, it is judged that the joining life is not practical and marked as x, if it is 144 hours or more and less than 288 hours, it is judged that there is no problem in practical use and marked as o, if it is 288 hours or more and less than 384 hours, it is judged that the joining life is excellent and marked as o, if it is 384 hours or more, it is judged that the joining life is particularly excellent and marked as o.

The crushed shape of the ball joint is evaluated by the ball joint obtained by joining the ball joints when viewed from directly above, and is determined based on the circularity thereof. The bonding target used an electrode in which an Al-0.5% Cu alloy film having a thickness of 1.0 μm was formed on a Si substrate. The observation was performed using an optical microscope, and 200 points were observed for 1 condition. The crushing shape of the ball joint was determined to be poor in the case of an oval shape having a large deviation from the perfect circle and in the case of having anisotropy in deformation. In the above evaluation, the "crushed shape" columns in tables 1 to 4 were all shown in the "crushed shape" columns, and the "crushed shape" column was all the excellent, and the "crushed shape" was all the excellent.

/>

(evaluation results)

The bonding wires according to examples 1 to 59 of the present invention had a Cu alloy core material and a Pd coating layer formed on the surface of the Cu alloy core material, and the thickness of the Pd coating layer was in the range of 0.015 to 0.150 μm as an appropriate range, and the FAB shapes were all good. Further, since these bonding wires contain at least 1 element selected from Ni, zn, rh, in, ir, pt and the concentration of the above elements is 0.011 to 2 mass% in total with respect to the entire wire, the high-temperature reliability of the ball bonding portion based on HTS evaluation is also good.

In the present invention, the reduction ratio at the time of wire drawing is 10% or more, and the heat treatment temperature in the heat treatment after wire drawing is 500 ℃ or less, so that the average crystal grain diameter of the core material cross section in the direction perpendicular to the axis of the bonding wire can be set to 0.9 to 1.5 μm in the result of measuring the crystal orientation of the core material cross section in the direction perpendicular to the axis of the bonding wire, in which the orientation ratio of < 100 > of the crystal orientation in the wire length direction with respect to the crystal orientation in the wire length direction is 15 degrees or less. As a result, although Ni, zn, rh, in, ir, pt is contained in the line, the resistance ratio (=maximum resistance/0.2% resistance) falls within the range of 1.1 to 1.6. Thus, the wedge bondability was good.

On the other hand, in comparative examples 4 to 7 and 12 to 14, the heat treatment temperature was set to a temperature of 600 ℃ or higher, so that the < 100 > orientation ratio in the longitudinal direction of the wire became less than 30%. In comparative examples 2, 6, 8, 9 and 14, the heat treatment temperature was set to a temperature of 620 ℃ or higher, so that the < 100 > orientation ratio in the longitudinal direction of the wire became less than 30%, and the average crystal grain diameter of the core material cross section became more than 1.5 μm. Thus, the comparative examples 2, 4 to 9 and 12 to 14 had poor or problematic wedge joining properties with a resistance ratio exceeding 1.6.

In comparative examples 1 and 3, the reduction ratio of the die was set to less than 10%, so that the average crystal grain size of the core material cross section was less than 0.9 μm, the endurance ratio was less than 1.1, and the wedge bonding properties were poor. In comparative examples 10 and 11, the < 100 > orientation ratio in the wire length direction became less than 30%, and the average crystal grain diameter of the core material cross section was less than 0.9 μm, both of which were poor in wedge bonding property. In comparative example 15, the average crystal grain size was 0.9 to 1.5. Mu.m, and the < 100 > orientation ratio in the wire length direction was 30% or more, but since the composition did not contain an element imparting connection reliability in a high-temperature environment, HTS, HAST, and wedge connectivity were poor. In comparative example 16, since no element imparting connection reliability in a high-temperature environment was contained, HTS and HAST were poor.

< examples 2-1 to 2-44 of the present invention >)

(sample)

First, a method for producing a sample will be described. Cu, which is a material of the core material, has a purity of 99.99 mass% or more, and the remainder is composed of unavoidable impurities. Ga. Ge, ni, ir, pt, pd, B, P, mg all use a material having a purity of 99 mass% or more and the balance of unavoidable impurities. Ga, ge, ni, ir, pt, pd, B, P, mg, which is an element added to the core material, is formulated so that the composition of the wire or core material becomes a target composition. The addition of Ga, ge, ni, ir, pt, pd, B, P, mg may be performed by simple substance, but in the case of an element having a high melting point as a simple substance and/or an addition amount of an extremely small amount, a Cu master alloy containing an addition element may be prepared in advance and blended so as to be a target addition amount.

The Cu alloy of the core material is manufactured by the following process: charging raw material into a cylindrical carbon crucible with diameter of phi 3-6 mm, using a high-frequency furnace, vacuum or N ₂ Heating to 1090-1300 ℃ under inert atmosphere such as Ar gas, melting, and cooling. The resulting alloy having a diameter of 3 to 6mm was drawn to a diameter of 0.9 to 1.2mm, and then successively drawn using a drawing die to produce a wire having a diameter of 300 to 600. Mu.m. A commercially available lubricant was used for drawing, and the drawing speed was set to 20 to 150 m/min. In order to remove the oxide film on the wire surface, acid washing treatment is performed with hydrochloric acid, and then a Pd coating layer having a thickness of 1 to 15 μm is formed so as to cover the entire surface of the Cu alloy of the core material. Further, a part of the wires forms an alloy skin layer containing Au and Pd with a thickness of 0.05 to 1.5 μm on the Pd coating layer. For the formation of the Pd coating layer and the alloy skin layer containing Au and Pd, an electrolytic plating method was used. The plating solution used was a commercially available semiconductor plating solution. Then, the heat treatment and the wire drawing process were repeated at 200 to 500℃to a diameter of 20. Mu.m. After processing, finally circulate N ₂ Or Ar gas is heat-treated so that the elongation at break becomes about 5 to 15%. The heat treatment method is to circulate N while continuously sweeping the wire ₂ Or Ar gas. The wire conveying speed is set to 20-200 m/min, the heat treatment temperature is 200-600 ℃, and the heat treatment time is set to 0.2-1.0 seconds.

The concentration analysis of the Pd coating layer and the alloy surface layer containing Au and Pd was performed by using an auger electron spectroscopy apparatus while sputtering Ar ions from the surface of the bonding wire in the depth direction. The thickness of the coating layer and the alloy skin layer is determined from the depth-wise concentration profile (depth unit is determined by SiO ₂ Conversion). A region having a Pd concentration of 50 atomic% or more and an Au concentration of less than 10 atomic% was used as the Pd coating layer, and a region having an Au concentration of 10 atomic% or more on the surface of the Pd coating layer was used as the alloy skin layer. Coating layer and alloy surfaceThe thickness of the skin layer and the Pd maximum concentration are reported in tables 5 and 6, respectively. The concentration of Pd in the Cu alloy core material was determined by the following method: the line section is exposed, and line analysis, dot analysis, and the like are performed by using an electron beam microscopic analyzer mounted in a scanning electron microscope. As a method for exposing the wire cross section, mechanical polishing, ion etching, or the like is used. The concentration of Ga, ge, ni, ir, pt, B, P, mg in the bonding wire is a liquid obtained by analyzing the bonding wire by a strong acid using an ICP emission spectrometry device or an ICP mass spectrometry device, and is detected as the concentration of the element contained in the entire bonding wire.

The structures of the respective samples produced by the above-described steps are shown in tables 5 and 6 below.

/>

(evaluation method)

The line surface was used as an observation surface, and the crystal structure was evaluated. As an evaluation method, an electron back scattering diffraction method (EBSD: electron Backscattered Diffraction) was used. The EBSD method is characterized in that it can observe the crystal orientation of the observation surface and can show the angle difference of the crystal orientation between adjacent measurement points, and can observe the crystal orientation with relatively ease and high accuracy even in a thin line such as a bonding wire.

When the EBSD method is performed with a curved surface such as a line surface as an object, attention is required. When a portion having a large curvature is measured, it is difficult to perform measurement with high accuracy. However, by fixing the bonding wire to be measured in a straight line on a plane and measuring the flat portion near the center of the bonding wire, measurement with high accuracy can be performed. Specifically, the measurement region may be as follows. The circumferential dimension is about the center of the longitudinal direction of the wire and is 50% or less of the wire diameter, and the longitudinal dimension of the wire is 100 μm or less. Preferably, the circumferential dimension is 40% or less of the wire diameter, and the longitudinal dimension is 40 μm or less, and if so, the measurement efficiency is improved by shortening the measurement time. In order to further improve the accuracy, it is preferable to measure 3 or more points to obtain average information in consideration of the deviation. The measurement positions may be spaced apart by 1mm or more to avoid proximity.

The orientation ratio of < 100 > of the crystal orientations in the longitudinal direction of the wire in the core material cross section in the direction perpendicular to the axis of the bonding wire, the crystal orientations having an angle difference of 15 degrees or less with respect to the longitudinal direction of the wire, and the average crystal grain size (μm) in the core material cross section in the direction perpendicular to the axis were obtained by the same methods as in examples 1 to 59 of the present invention. The 0.2% endurance and the maximum endurance were evaluated in the same manner as in inventive examples 1 to 59, and the endurance ratio was calculated from the above formula (1).

The bonding reliability of the ball bonding portion in a high-temperature and high-humidity environment or a high-temperature environment was determined by preparing a sample for evaluating bonding reliability, performing HAST and HTS evaluation, and determining the bonding life of the ball bonding portion in each test. The sample for evaluating the bonding reliability was prepared by forming an Al-1.0% Si-0.5% Cu alloy film having a thickness of 0.8 μm on a Si substrate on a general metal frame, ball bonding the film by a commercially available wire bonder, and sealing the film with a commercially available epoxy resin. The ball circulates N at a flow rate of 0.4 to 0.6L/min ₂ +5％H ₂ The gas is formed while having a size in the range of from 33 to 34. Mu.m.

For HAST evaluation, a sample for evaluating the bonding reliability was exposed to a high-temperature and high-humidity environment having a temperature of 130 ℃ and a relative humidity of 85% using an unsaturated pressure cooker retort tester, and a bias voltage of 7V was applied. The joint life of the ball joint was set to a time period of 1/2 of the initial shear strength obtained by conducting the shear test of the ball joint every 48 hours. The shear test after the high temperature and high humidity test is performed after the resin is removed by acid treatment to expose the ball joint.

As a shear tester for HAST evaluation, a tester manufactured by DAGE company was used. The shear strength values were averaged over the measured values at 10 points of the randomly selected ball joint. In the above evaluation, if the joining life is less than 96 hours, it is judged that there is a problem in practice and marked as x, if it is 96 hours or more and less than 144 hours, it is judged that there is a problem in practice and marked as Δ, if it is 144 hours or more and less than 288 hours, it is judged that there is no problem in practice and marked as o, if it is 288 hours or more, it is judged that it is particularly excellent and marked as o, and both are described in the "HAST" columns of tables 5 and 6.

A shear tester manufactured by DAGE company was used for the HTS evaluation. The shear strength values were averaged over the measured values at 10 points of the randomly selected ball joint. In the above evaluation, if the joining life is 500 hours or more and less than 1000 hours, it is judged that the joining life is practical but needs improvement, and if it is 1000 hours or more and less than 3000 hours, it is judged that the joining life is practically no problem, and if it is 3000 hours or more, it is judged that the joining life is particularly excellent, and it is marked as excellent.

The ball formability (FAB shape) was evaluated by observing the ball before joining, and judging whether or not there was air bubbles on the ball surface and deformation of the ball which was originally a sphere. If any of the above-described phenomena occurs, the failure is determined. In order to suppress oxidation in the melting step, N was blown at a flow rate of 0.5L/min for the formation of balls ₂ The gas is carried out at the same time. The size of the spheres was set to 34 μm. 50 balls were observed for 1 condition. SEM was used for observation. In the evaluation of the sphericity, if 5 or more failures occurred, the failure was judged to be a problem and marked as xIf the number of defects is 3 to 4, it is judged that the method is practical but the method has a slight problem and is marked as delta, if the number of defects is 1 to 2, it is judged that the method has no problem and is marked as o, and if the method has no defects, it is judged that the method has excellent and is marked as o, and the results are shown in the FAB shape columns of tables 5 and 6.

The wedge bondability in the wire bonding portion was evaluated by bonding 1000 wires to the lead portion of the lead frame, and the evaluation was determined based on the occurrence frequency of peeling of the bonding portion. The lead frame used was an Fe-42 at% Ni alloy lead frame to which an Ag-plated layer having a thickness of 1 to 3 μm was applied. In this evaluation, the stage temperature was set to 150 ℃. In the above evaluation, if 11 or more failures occurred, it was judged that the failure was marked as x, and if the failure was 6 to 10, it was judged that the failure was practical but that the failure was slightly problematic, and if the failure was 1 to 5, it was judged that the failure was no problem, and if the failure was not occurred, it was judged that the failure was excellent, and the failure was marked as excellent, and the "wedge-bonding property" columns of tables 5 and 6 are both described.

The crushed shape of the ball joint is evaluated by the ball joint obtained by joining the ball joints when viewed from directly above, and is determined based on the circularity thereof. The bonding target used an electrode in which an Al-0.5% Cu alloy film having a thickness of 1.0 μm was formed on a Si substrate. The observation was performed using an optical microscope, and 200 points were observed for 1 condition. The crushing shape of the ball joint was determined to be poor in the case of an oval shape having a large deviation from the perfect circle and in the case of having anisotropy in deformation. In the above evaluation, if 6 or more failures occurred, it was judged that the failure was marked as "x", if the failure was 4 to 5, it was judged that the failure was practical, but if the failure was slightly problematic, it was judged that the failure was no problem, if the failure was 1 to 3, it was judged that the failure was marked as "o", and if all of the degrees of circularity were good, it was judged that the failure was particularly excellent, and the failure was marked as "excellent", which are shown in the column of "crushed shapes" in tables 5 and 6.

[ Tilt ]

100 lead frames for evaluation were bonded to each other with a loop length of 5mm and a loop height of 0.5 mm. As an evaluation method, the line upright portion was observed from the chip horizontal direction, and the evaluation was performed with the interval (tilt interval) when the interval between the vertical line passing through the center of the ball joint portion and the line upright portion was the largest. If the pitch interval is smaller than the wire diameter, the pitch is good, and if the pitch interval is larger than the wire diameter, the standing part is inclined, so that the inclination is judged to be poor. 100 bonded wires were observed with an optical microscope, and the number of defective inclinations was counted. If 7 or more failures occurred, it was judged that the failure was marked as "x", and if the failure was 4 to 6, it was judged that the failure was practical, but if the failure was slightly problematic, it was judged that the failure was no problem, and if the failure was 1 to 3, it was judged that the failure was excellent, and if the failure was not occurred, it was marked as "excellent", and the "tilt" columns of tables 5 and 6 are both described.

(evaluation results)

As shown in tables 5 and 6, the bonding wires according to examples 2-1 to 2-44 of the present invention include a Cu alloy core material and a Pd coating layer formed on the surface of the Cu alloy core material, and the bonding wire includes 1 or more elements selected from Ga and Ge, and the total concentration of the elements is 0.011 to 1.5 mass% with respect to the entire wire. From this, it was confirmed that the bonding wires according to examples 2-1 to 2-44 of the present invention were able to obtain ball bonding portion reliability in a HAST test under a high-temperature and high-humidity environment having a temperature of 130 ℃ and a relative humidity of 85%.

Regarding the present invention example having an alloy skin layer containing Au and Pd on the Pd coating layer, it was confirmed that: the alloy skin layer containing Au and Pd has a thickness of 0.0005 to 0.050 μm, thereby obtaining excellent wedge bonding properties.

The bonding wire according to the present invention, which further includes at least 1 element selected from Ni, ir, pt, pd, was found to have more excellent high-temperature reliability of the ball bonding portion based on HTS evaluation.

In the present invention, the bonding wire further includes at least 1 element selected from B, P, mg, and the concentration of each element is 1 to 200 mass ppm with respect to the entire wire, whereby the crushed shape of the ball bonding portion is good.

< examples 3-1 to 3-50 of the present invention >)

(sample)

First, a method for producing a sample will be described. Cu, which is a material of the core material, has a purity of 99.99 mass% or more, and the remainder is composed of unavoidable impurities. As, te, sn, sb, bi, se, ni, zn, rh, in, ir, pt, ga, ge, pd, B, P, mg, ca, la a material having a purity of 99 mass% or more and the balance of unavoidable impurities is used. As, te, sn, sb, bi, se, ni, zn, rh, in, ir, pt, ga, ge, pd, B, P, mg, ca, la, which is an element added to the core material, is formulated so that the composition of the wire or core material becomes a target composition. The addition of As, te, sn, sb, bi, se, ni, zn, rh, in, ir, pt, ga, ge, pd, B, P, mg, ca, la may be performed by simple substance, but in the case where the element is a simple substance having a high melting point and/or the addition amount is an extremely small amount, a Cu master alloy containing the addition element may be prepared in advance and blended so as to be a target addition amount.

The Cu alloy of the core material is manufactured by the following process: charging raw material into a cylindrical carbon crucible with diameter of phi 3-6 mm, using a high-frequency furnace, vacuum or N ₂ Heating to 1090-1300 ℃ under inert atmosphere such as Ar gas, melting, and cooling. The resulting alloy having a diameter of 3 to 6mm was drawn to a diameter of 0.9 to 1.2mm, and then successively drawn using a drawing die to produce a wire having a diameter of 300 to 600. Mu.m. A commercially available lubricant was used for drawing, and the drawing speed was set to 20 to 150 m/min. In order to remove the oxide film on the wire surface, acid washing treatment is performed with hydrochloric acid, and then a Pd coating layer having a thickness of 1 to 15 μm is formed so as to cover the entire surface of the Cu alloy of the core material. Further, a part of the wires forms an alloy skin layer containing Au and Pd with a thickness of 0.05 to 1.5 μm on the Pd coating layer. For the formation of the Pd coating layer and the alloy skin layer containing Au and Pd, an electrolytic plating method was used. The plating solution used was a commercially available semiconductor plating solution. Then, the heat treatment and the wire drawing process were repeated at 200 to 500℃to a diameter of 20. Mu.m. After processing, finally circulate N ₂ Or Ar gas is heat-treated so that the elongation at break becomes about 5 to 15%. Heat treatment formulaThe method comprises passing N through a line while continuously sweeping the line ₂ Or Ar gas. The wire conveying speed is set to 20-200 m/min, the heat treatment temperature is 200-600 ℃, and the heat treatment time is set to 0.2-1.0 seconds.

The concentration analysis of the Pd coating layer and the alloy surface layer containing Au and Pd was performed by auger electron spectroscopy while cutting the surface of the bonding wire in the depth direction by sputtering or the like. From the obtained depth-direction concentration profile, the thickness of the Pd coating layer, the thickness of the alloy skin layer containing Au and Pd, and the Pd maximum concentration were obtained.

Examples 3-1 to 3-50 of the present invention contained an element selected from As, te, sn, sb, bi, se in the core material.

In the present invention examples 3-34 to 3-44, cu was present on the outermost surface of the bonding wire. Therefore, the column of "line surface Cu concentration" is set in table 7, and the results of measuring the surface of the bonding wire by the auger electron spectroscopy apparatus are described. The heat treatment temperature and time of the bonding wire are selected so that the outermost surface contains Cu at a predetermined concentration. In the present invention examples 3-1 to 3-33 and 3-45 to 3-50, the conditions for heat treatment were set so that Cu was not present on the outermost surface, and Cu was not detected even by using an Auger electron spectroscopy apparatus.

The structures of the respective samples produced by the above-described steps are shown in tables 7 and 8.

/>

(evaluation method)

The orientation ratio of < 100 > of the crystal orientations in the line length direction among the crystal orientations in the line length direction in the core material cross section in the direction perpendicular to the axis of the bonding line, the crystal orientation angle difference of 15 degrees or less with respect to the line length direction, and the average crystal grain diameter (μm) of the core material cross section in the direction perpendicular to the axis were obtained by the same methods as in the present invention examples 1 to 59. The 0.2% endurance and the maximum endurance were evaluated in the same manner as in inventive examples 1 to 59, and the endurance ratio was calculated from the above formula (1).

The bonding reliability of the ball bonding portion in a high-temperature and high-humidity environment or a high-temperature environment was determined by preparing a sample for evaluating bonding reliability, performing HAST and HTS evaluation, and determining the bonding life of the ball bonding portion in each test. The sample for evaluating the bonding reliability was prepared by forming an Al-1.0% Si-0.5% Cu alloy film having a thickness of 0.8 μm on a Si substrate on a general metal frame, ball bonding the film with a commercially available wire bonder, and packaging the film with a commercially available epoxy resin. The ball circulates N at a flow rate of 0.4 to 0.6L/min ₂ +5％H ₂ The formation of the gas is carried out at the same time,the size is set to be in the range of phi 33-34 mu m.

For HAST evaluation, a sample for evaluating the bonding reliability was exposed to a high-temperature and high-humidity environment having a temperature of 130 ℃ and a relative humidity of 85% using an unsaturated pressure cooker retort tester, and a bias voltage of 5V was applied. The joint life of the ball joint was set to a time period of 1/2 of the initial shear strength obtained by conducting the shear test of the ball joint every 48 hours. The shear test after the high temperature and high humidity test is performed after the resin is removed by acid treatment to expose the ball joint.

As a shear tester for HAST evaluation, a tester manufactured by DAGE company was used. The shear strength values were averaged over the measured values at 10 points of the randomly selected ball joint. In the above evaluation, if the joining life is less than 96 hours, it is judged that there is a problem in practice and marked as x, if it is 96 hours or more and less than 144 hours, it is judged that there is a problem in practice and marked as Δ, if it is 144 hours or more and less than 288 hours, it is judged that there is no problem in practice and marked as o, if it is 288 hours or more and less than 384 hours, it is judged that there is excellent and marked as o, if it is 384 hours or more, it is judged that there is particularly excellent and marked as o.

A shear tester manufactured by DAGE company was used for the HTS evaluation. The shear strength values were averaged over the measured values at 10 points of the randomly selected ball joint. In the above evaluation, if the joining life was 500 hours or more and less than 1000 hours, it was judged that the joining life was practical but improvement was required, and if it was 1000 hours or more and less than 3000 hours, it was judged that the joining life was practically no problem, and if it was 3000 hours or more, it was particularly excellent, and it was marked that the joining life was excellent, and the joining life was marked as "delta", and the "HTS" columns in tables 7 and 8 were all described.

The ball formability (FAB shape) was evaluated by observing the ball before joining, and judging whether or not there was air bubbles on the ball surface and deformation of the ball which was originally a sphere. If any of the above-described phenomena occurs, the failure is determined. In order to suppress oxidation in the melting step, N was blown at a flow rate of 0.5L/min for the formation of balls ₂ The gas is carried out at the same time. The size of the spheres was set to 34 μm. 50 balls were observed for 1 condition. SEM was used for observation. In the evaluation of the sphere formation, if 5 or more failures occurred, it was judged that there was a problem and marked as x, if 3 to 4 failures were found, it was judged that the sphere was practical but slightly problematic and marked as Δ, if 1 to 2 failures were found, it was judged that there was no problem and marked as o, and if no failure occurred, it was judged that the sphere was excellent and marked as o, and the "FAB shape" columns in tables 7 and 8 were both described.

The wedge bondability in the wire bonding portion was evaluated by bonding 1000 wires to the lead portion of the lead frame, and the evaluation was determined based on the occurrence frequency of peeling of the bonding portion. The lead frame used was an Fe-42 at% Ni alloy lead frame to which an Ag-plated layer of 1 to 3 μm was applied. In this evaluation, the stage temperature was set to 150 ℃. In the above evaluation, when 11 or more failures occurred, it was judged that the failure was marked as x, and if the failure was 6 to 10, it was judged that the failure was practical but that the failure was slightly problematic, and when the failure was 1 to 5, it was judged that the failure was no problem, and when the failure was not occurred, it was judged that the failure was excellent, and the failure was marked as excellent, and the "wedge-bonding property" columns of tables 7 and 8 were both described.

The crushed shape of the ball joint is evaluated by the ball joint obtained by joining the ball joints when viewed from directly above, and is determined based on the circularity thereof. The bonding target used an electrode in which an Al-0.5% Cu alloy film having a thickness of 1.0 μm was formed on a Si substrate. The observation was performed using an optical microscope, and 200 points were observed for 1 condition. The crushing shape of the ball joint was determined to be poor in the case of an oval shape having a large deviation from the perfect circle and in the case of having anisotropy in deformation. In the above evaluation, if 6 or more failures occurred, it was judged that the failure was marked as "x", if the failure was 4 to 5, it was judged that the failure was practical, but if the failure was slightly problematic, it was judged that the failure was no problem, if the failure was 1 to 3, it was judged that the failure was marked as "o", and if all of the degrees of circularity were good, it was judged that the failure was particularly excellent, and the failure was marked as "excellent", which are shown in the column of "crushed shape" in tables 7 and 8.

[ Tilt ]

100 lead frames for evaluation were bonded to each other with a loop length of 5mm and a loop height of 0.5 mm. As an evaluation method, the line upright portion was observed from the chip horizontal direction, and the evaluation was performed with the interval (tilt interval) when the interval between the vertical line passing through the center of the ball joint portion and the line upright portion was the largest. If the pitch interval is smaller than the wire diameter, the pitch is good, and if the pitch interval is larger than the wire diameter, the standing part is inclined, so that the inclination is judged to be poor. 100 bonded wires were observed with an optical microscope, and the number of defective inclinations was counted. If 7 or more failures occurred, it was judged that the failure was marked as "x", and if the failure was 4 to 6, it was judged that the failure was practical, but if the failure was slightly problematic, it was judged that the failure was no problem, and if the failure was 1 to 3, it was judged that the failure was excellent, and if the failure was not occurred, it was marked as "excellent", and the "tilt" columns of tables 7 and 8 were both described.

(evaluation results)

The bonding wire according to examples 3-1 to 3-50 of the present invention has a Cu alloy core material and a Pd coating layer formed on the surface of the Cu alloy core material, and the bonding wire contains at least 1 element selected from As, te, sn, sb, bi, se, and the concentration of the above elements is 0.1 to 100 mass ppm in total with respect to the entire wire. Thus, it was confirmed that the bonding wires according to examples 3-1 to 3-50 of the present invention were capable of obtaining ball bond reliability in a HAST test under a high-temperature and high-humidity environment at 130 ℃ and a relative humidity of 85%.

Regarding the present invention example having an alloy skin layer containing Au and Pd on the Pd coating layer, it was confirmed that: the alloy skin layer containing Au and Pd has a layer thickness of 0.0005 to 0.050 μm, thereby obtaining excellent wedge bonding properties.

Regarding the present invention examples 3-21 to 3-33, 3-35, 3-37, 3-39 to 3-44, it was confirmed that: the bonding wire further includes at least 1 element selected from Ni, zn, rh, in, ir, pt, ga, ge, and the concentrations of the elements are 0.011 to 1.2 mass% with respect to the entire wire, and the concentration of Pd contained in the Cu alloy core material is 0.05 to 1.2 mass%, whereby the high-temperature reliability of the ball bonding portion based on HTS evaluation is good.

In the present invention examples 3-22 to 3-26 and 3-29 to 3-32, at least 1 element selected from B, P, mg, ca, la was further contained in the bonding wire, and the concentrations of the above elements were 1 to 100 mass ppm with respect to the entire wire, whereby the FAB shape was good and the wedge bonding property was good.

With respect to the present invention examples 3-34 to 3-44, the wire contained As, te, sn, sb, bi, se, and Cu was present at the outermost surface of the wire. Thus, regarding the present invention examples 3-34 to 3-44, the results of the hast evaluation were excellent or very excellent, and the effect of Cu being present on the outermost surface was observed.

< examples 4-1 to 4-15 of the present invention >)

As a raw material of the bonding wire, cu having a purity of 99.99 mass% or more was prepared as an additive element for producing the Cu alloy core material, ni, pd, pt, au, P, B, be, fe, mg, ti, zn, ag, si was prepared, pd having a purity of 99.99 mass% or more was prepared for forming the coating layer, and Au having a purity of 99.99 mass% or more was prepared for forming the alloy skin layer. After weighing Cu and an additive element raw material as starting materials, the materials were heated and melted under high vacuum to obtain copper alloy ingots having a diameter of about 10 mm. Then, this ingot was forged, rolled, and wire-drawn to prepare a Cu alloy wire having a diameter of 500. Mu.m. Next, a Pd coating layer having a thickness of 1 to 3 μm was applied to the surface of the Cu alloy wire by electrolytic plating, and an Au skin layer having a thickness of 0.05 to 0.2 μm was applied to the surface of the coating layer, thereby obtaining a multilayer wire. The final thicknesses of the Pd coating layer and the AuPd alloy skin layer are shown in table 8. Here, the boundary between the core material and the coating layer was set at a position where the Pd concentration was 50 at%, and the boundary between the coating layer and the alloy skin layer was set at a position where the Au concentration was 10 at%. Then, continuous wire drawing was performed at a wire drawing speed of 100 to 700m/min and a die reduction of 8 to 30%, so that the final wire diameters described in Table 8 were obtained. The thickness of the alloy skin layer, the Au maximum concentration, the surface Cu concentration, and the thickness of the coating layer are controlled by performing heat treatment 2 to 3 times between wiredrawing processes. The conditions at this time are that the temperature is 500 to 700℃and the speed is 10 to 70m/min when the wire diameter is 200 to 250. Mu.m, that the temperature is 450 to 650℃and the speed is 20 to 90m/min when the wire diameter is 70 to 100. Mu.m, and that the temperature is 300 to 500℃and the speed is 30 to 100m/min when the final wire diameter is small and the wire diameter is 40 to 70. Mu.m. Then, heat treatment was performed at a final wire diameter under the conditions of the temperature and the speed of 30 to 120m/min in Table 8. In order to diffuse Cu to the surface, the oxygen concentration in the heat treatment furnace is set to be 0.2 to 0.7% higher than usual when the heat treatment is performed only 1 time. This heat treatment is preferable in the end if possible because oxidation of Cu is easily caused when drawing is repeatedly performed after Cu appears on the surface. In the heat treatment other than this, the oxygen concentration in the heat treatment furnace is made to be less than 0.2%, whereby excessive oxidation of the alloy skin layer is suppressed, and stable thickness, composition, and the like are controlled. Thus, a bonding wire having a diameter of 15 to 25 μm was obtained.

Concentration analysis of the coating layer, the alloy skin layer, and concentration analysis of Ni, pd, pt, au in the Cu alloy core material were performed by using an AES apparatus while sputtering Ar ions from the surface of the bonding wire in the depth direction. The thickness of the coating layer and the alloy skin layer were determined from the depth-wise concentration profile (depth unit is determined by SiO ₂ Conversion). In observation of element distribution, analysis is also performed by using EPMA, EDX apparatus, or the like. A region having a Pd concentration of 50 atomic% or more and an Au concentration of less than 10 atomic% was used as the coating layer, and a region having an Au concentration in the range of 10 atomic% or more on the surface of the coating layer was used as the alloy skin layer. Table 8 shows the coating layer and the alloy skin layerThickness and composition. The concentration of P, B, be, fe, mg, ti, zn, ag, si in the bonding wire was measured by an ICP emission spectrometry device, an ICP mass spectrometry device, or the like. The orientation ratio of < 100 > of the crystal orientations in the line length direction among the crystal orientations in the core material cross section in the direction perpendicular to the axis of the bonding line, the crystal orientations having an angle difference of 15 degrees or less with respect to the line length direction, and the average crystal grain diameter (μm) of the core material cross section in the direction perpendicular to the axis were obtained by the same methods as in examples 1 to 59 of the present invention. The 0.2% endurance and the maximum endurance were evaluated by the same methods as in inventive examples 1 to 59, and the endurance ratio was calculated from the above formula (1).

For the connection of the bonding wires, a commercially available automatic wire bonding machine is used. Immediately before bonding, a ball was made at the tip of the bonding wire by arc discharge, but its diameter was selected so as to be 1.7 times the diameter of the bonding wire. The atmosphere at the time of ball production was nitrogen.

As the bonding targets of the bonding wires, al electrodes having a thickness of 1 μm formed on Si chips and leads of a lead frame having a Pd-plated surface were used, respectively. After ball bonding is performed between the ball portion thus manufactured and the electrode heated to 260 ℃, the bus bar portion of the bonding wire and the wire heated to 260 ℃ are wedge bonded, and the ball portion is manufactured again, whereby bonding is continuously repeated. The loop length was set to be both 3mm and 5mm, and the loop height was set to be both 0.3mm and 0.5 mm.

The bonding wire was evaluated for wedge bonding property, and fish tail symmetry. Regarding bondability, bonded portions of 100 bonding wires in a wedge bonded state were observed, and the wire from which the bonded portions were peeled was counted and denoted as NG. Regarding the symmetry of the fish tail, the joints of the 100 bonding wires in the wedge-bonded state were observed, and the symmetry was evaluated. The length from the center to the left end and the length from the center to the right end of the fishtail crimp were measured, and the line with a difference of 10% or more was counted and recorded as NG. Regarding the bondability and the fish tail symmetry, the case where NG is 0 was marked as excellent, the case where NG is 1 to 10 was marked as o, and the case where NG is 11 or more was marked as x.

Regarding 1st bondability (ball bondability) of the bonding wire, evaluation was made for HTS test, HAST test, and FAB shape. Regarding the HTS test, the same methods as those of examples 1 to 59 of the present invention were used for evaluation. In order to evaluate the soundness of the ball joint in the HAST test, the bonded semiconductor device was placed in a high-temperature and high-humidity furnace at a temperature of 130 ℃ and a relative humidity of 85% rh (Relative Humidity Humidity) and 5V, and taken out every 48 hours to evaluate the bonded semiconductor device. As an evaluation method, the resistance was measured, and the sample with the increased resistance was denoted as NG. The time until NG was 480 hours or more was marked as excellent, the time at least 384 hours and less than 480 hours was marked as o, and the time at less than 384 hours was marked as x.

Regarding the FAB shapes, 100 FABs were fabricated on a lead frame, and observed with SEM. The spherical shape was designated as OK, the core shift and shrinkage cavity were designated as NG, and the number thereof was counted. Regarding the FAB shape, 0 NG was marked as "good", 1 to 5 NG was marked as "good", 6 to 10 NG was marked as "delta", and 11 or more NG was marked as "X". The sum of ∈and ∈was acceptable, and the delta was acceptable but slightly inferior. The concentrations (mass%) of Ni, pd and Pt in Table 9 represent the concentrations in the Cu alloy core material.

In the present invention examples 4-1 to 4-15, the Cu alloy core material contained 0.1 to 3.0 mass% in total of 1 or more elements (metal elements of group 10 of the periodic table) selected from Ni, pd and Rt, and the Cu concentration at the outermost surface of the bonding wire was 1 to 10 atomic%. Thus, in the present invention examples 4-1 to 4-15, the wedge joint was excellent in the joint property and the symmetry of the fish tail, and also excellent in HTS, FAB shape and HAST.

Claims

1. A bonding wire for a semiconductor device, characterized by comprising a Cu alloy core material and a Pd coating layer formed on the surface of the Cu alloy core material,

in the result of measuring the crystal orientation of the core material section in the direction perpendicular to the axis of the bonding wire, the orientation ratio of < 100 > of the crystal orientation of 15 degrees or less with respect to the angle difference in the longitudinal direction of the wire is 30% or more,

the average crystal grain diameter of the core material section in the direction perpendicular to the axis of the bonding wire is 0.9 μm or more and 1.5 μm or less,

the bonding wire contains 1 or more elements selected from Co, rh, ir, ni, pd, pt, ag, au, zn, al, in, sn, P, as, sb, bi, se, te, and the concentration of the elements relative to the entire wire is 2.0 mass% or less in total.

2. The bonding wire for semiconductor device according to claim 1, wherein,

the concentration of the element is 0.011 mass% or more in total with respect to the entire line.

3. The bonding wire for semiconductor device according to claim 1, wherein,

the concentration of the element is 0.050 mass% or more in total with respect to the whole line.

4. The bonding wire for semiconductor device according to claim 1, wherein,

the concentration of the element is 0.070 mass% or more in total with respect to the whole line.

5. The bonding wire for semiconductor device according to claim 1, wherein,

the concentration of the element is 0.10 mass% or more in total with respect to the whole line.

6. The bonding wire for semiconductor device according to claim 1, wherein,

the concentration of the element is 0.20 mass% or more in total with respect to the whole line.

7. The bonding wire for semiconductor device according to claim 1, wherein,

the concentration of the element is 1.8 mass% or less in total with respect to the entire line.

8. The bonding wire for semiconductor device according to claim 1, wherein,

the resistance ratio defined by the following formula (1) is 1.1 to 1.6,

Resistance ratio = maximum resistance/0.2% resistance (1).

9. The bonding wire for semiconductor device according to claim 1, wherein,

the Pd coating layer has a thickness of 0.015 μm or more and 0.150 μm or less.

10. The bonding wire for semiconductor device according to claim 9, wherein,

the Pd coating layer has a thickness of 0.02 μm or more.

11. The bonding wire for semiconductor device according to claim 9, wherein,

the Pd coating layer has a thickness of 0.03 μm or more.

12. The bonding wire for semiconductor device according to claim 9, wherein,

the Pd coating layer has a thickness of 0.100 μm or less.

13. The bonding wire for semiconductor device according to claim 1, wherein,

the Pd coating layer also has an alloy surface layer containing Au and Pd.

14. The bonding wire for semiconductor device according to claim 13, wherein,

the thickness of the alloy skin layer containing Au and Pd is below 0.050 μm.

15. The bonding wire for semiconductor device according to claim 14, wherein,

the thickness of the alloy skin layer containing Au and Pd is more than 0.001 μm.

16. The bonding wire for semiconductor device according to claim 14, wherein,

The thickness of the alloy skin layer containing Au and Pd is more than 0.002 μm.

17. The bonding wire for semiconductor device according to claim 14, wherein,

the thickness of the alloy skin layer containing Au and Pd is 0.030 μm or less.

18. The bonding wire for semiconductor device according to claim 1, wherein,

the bonding wire includes at least 1 or more elements selected from Ni, zn, rh, in, ir, pt, and the concentration of the elements is 0.011 mass% or more and 2 mass% or less in total with respect to the entire wire.

19. The bonding wire for semiconductor device according to claim 18, wherein,

20. The bonding wire for semiconductor device according to claim 18, wherein,

21. The bonding wire for semiconductor device according to claim 18, wherein,

22. The bonding wire for semiconductor device according to claim 18, wherein,

23. The bonding wire for semiconductor device according to claim 1, wherein,

the bonding wire contains 1 or more elements selected from As, te, sn, sb, bi, se, and the concentration of the elements is 0.1 mass ppm or more and 100 mass ppm or less in total, sn 10 mass ppm or less, sb 10 mass ppm or less, and Bi 1 mass ppm or less based on the entire wire.

24. The bonding wire for semiconductor device according to claim 23, wherein,

the concentration of the elements is 1 mass ppm or more in total with respect to the whole line.

25. The bonding wire for semiconductor device according to claim 23, wherein,

the concentration of the elements is 2 mass ppm or more in total with respect to the whole line.

26. The bonding wire for semiconductor device according to claim 23, wherein,

the concentration of the elements is 3 mass ppm or more in total with respect to the whole line.

27. The bonding wire for semiconductor device according to claim 1, wherein,

the bonding wire further includes at least 1 or more elements selected from B, P, mg, ca, la, and the concentration of the elements is 1 mass ppm or more and 200 mass ppm or less, respectively, with respect to the entire wire.

28. The bonding wire for semiconductor device according to claim 27, wherein,

the concentrations of the elements are 5 mass ppm or more, respectively, relative to the entire line.

29. The bonding wire for semiconductor device according to claim 1, wherein,

cu is present at the outermost surface of the bond wire.

30. The bonding wire for semiconductor device according to any one of claims 1 to 29, wherein,

the Cu alloy core material contains 0.1 to 3.0 mass% of a metal element of group 10 of the periodic table, and the Cu concentration at the line outermost surface is 1 atomic% or more.

31. The bonding wire for semiconductor device according to claim 30, wherein,

the concentration of the metal elements of group 10 of the periodic table in the Cu alloy core material is 0.5 mass% or more in total.

32. The bonding wire for semiconductor device according to claim 30, wherein,

the Cu alloy core material contains Ni as a metal element of group 10 of the periodic table.