Accelerated testing method for reliability of intermetallic compound in wire bonding

Abstract

The invention relates to the technical field of chip packaging, in particular to an accelerated testing method for reliability of an intermetallic compound in wire bonding. Comprising the following steps: S1, carrying out a high-temperature test; s2, measuring the resistance value of each sample pin; s3, setting failure conditions; s4, calculating the cumulative failure rate when each sample fails; s5, arranging the test data into a table; s6, drawing in a logarithmic normal distribution coordinate system and acquiring a fitting linear equation of each acceleration temperature; s7, performing linear fitting; s8, calculating the time when the cumulative failure rate is 50% at each acceleration temperature; s9, drawing a mutual relation between the two; s10, performing linear fitting on each data point again to obtain a straight slope; and S11, substituting into an Arrhenius equation to obtain an acceleration coefficient at the acceleration test temperature so as to obtain the test time at the acceleration test temperature. Compared with the prior art, the testing time of the reliability of the intermetallic compound in lead bonding is shortened, and the time and the cost are saved.

Description

Technical Field

[0001] The invention relates to the technical field of chip packaging, in particular to an accelerated testing method for reliability of intermetallic compounds in wire bonding. Background Art

[0002] The wire bonding process is a very important process in the semiconductor packaging process. It usually refers to bonding the chip in the package to the lead frame or substrate with a metal wire under the pressure of the bonding tool splitter and ultrasonic vibration at a certain temperature to achieve electrical interconnection between the chip and the peripheral circuit. The most commonly used metal wires are gold wires, silver wires and copper wires, as well as a series of bonding wire materials formed on this basis by doping, alloying, compounding or surface treatment. When the solder ball is formed, a small amount of intermetallic compounds (IMC) is formed in the early stage. A small amount of intermetallic compounds has a certain strengthening effect on the interface bonding strength. However, with the increase of usage time and the increase of temperature, the amount of intermetallic compounds continues to increase. When there are too many intermetallic compounds, the bonding strength is reduced, it becomes brittle, and the contact resistance increases. The brittle intermetallic compounds will cause fatigue damage to the bonding point when subjected to periodic stress, which may eventually lead to device open circuit or degradation of the device's electrical performance.

[0003] The reliability test of intermetallic compounds in wire bonding is usually carried out according to the high temperature storage life test condition C defined by the Joint Electron Device Engineering Council (JEDEC) of the United States, that is, 150℃ / 1000 hours. The semiconductor packaging industry has the characteristics of large output, short cycle, high process stability requirements, and the need to quickly locate abnormal points for process improvement. The 1000-hour test time is often unacceptable for the need for rapid DOE (Design of Experiment) verification to locate abnormal points and improve processes in a short period of time and quickly predict product life. In order to solve this problem, a new technology needs to be developed to accelerate the test time of intermetallic compound reliability in wire bonding to save time and cost. Summary of the invention

[0004] In order to overcome the deficiencies of the prior art, the present invention provides an accelerated testing method for the reliability of intermetallic compounds in wire bonding.

[0005] To achieve the above object, a method for accelerating the reliability test of intermetallic compounds in wire bonding is designed, which is characterized by comprising the following steps: S1, select multiple samples and multiple accelerated high temperature temperatures, and conduct high temperature tests respectively; S2, measure the resistance value of each sample pin under different high temperature environments; S3, setting the failure condition, and the time when the failure condition is reached is called the failure time; S4, calculate the cumulative failure rate when each sample fails; S5, arranging the test data into a table, the test data including temperature, product failure time, and cumulative failure rate; S6, obtain the distribution of the failure time of each sample to the cumulative failure rate of the product, plot the data in the test data table in a log-normal distribution coordinate system and obtain the fitted linear equation of each accelerated temperature, the fitted linear equation is y=ax+b, where y is the cumulative failure rate, x is the logarithmic value of the failure time, a is the slope, and b is the intercept; S7, perform linear fitting on the data points of each acceleration temperature to confirm whether the slopes of each straight line are similar. If so, proceed to step S8; if not, lower the selected value of the highest temperature and adjust the interval value of each temperature to repeat steps S1 to S7; S8, according to the linear equation, calculate the time when the cumulative failure rate is 50% at each accelerated temperature, that is, the median life t(0.5); S9, plot the logarithm of the median life span, i.e., Lnt(0.5), and the inverse of the temperature value in the Arrhenius equation linear coordinate system to show the relationship between the two; S10, linear fitting is performed again on each data point to obtain the slope of the straight line, which is the activation energy value of the failure mechanism of the intermetallic compound. The fitted linear equation is m=cn+d, where m is Lnt(0.5), n is (1 / T)*10000, c is the slope, that is, the activation energy value of the failure mechanism, and d is the intercept; S11, substitute the activation energy value, accelerated test temperature, and normal test temperature into the Arrhenius equation to obtain the acceleration coefficient at the accelerated test temperature, thereby obtaining the test time at the accelerated test temperature; In step S11, the method for obtaining the acceleration coefficient at the accelerated test temperature according to the Arrhenius equation is as follows: AF=L1 / L2=exp[Ea / k (1 / T1-1 / T2) ], where AF is the acceleration coefficient, L1 is the life under normal conditions, L2 is the life under accelerated conditions, Ea is the activation energy value of the failure mechanism, and k is the Boltzmann constant (8.62X10 -5 ), T1 is the conventional test temperature, T2 is the accelerated test temperature; The method for obtaining the test time at the accelerated test temperature based on the acceleration coefficient is as follows: t1=t2 / AF, where t1 is the test time at the accelerated test temperature, t2 is the test time at the conventional test temperature, and AF is the acceleration coefficient.

[0006] In the step S1, the temperature of the accelerated high temperature is 180-300°C.

[0007] In the step S2, the resistance value is measured once every hour.

[0008] In step S9, the unit of the temperature value is Kelvin.

[0009] In the step S3, the failure condition is the resistance change rate (Ri-R0) / R0 ≥ 10%; wherein R0 is the initial value of the resistance value, and Ri (i=1,2...) is the resistance value tested in the i-th hour.

[0010] In step S4, the cumulative failure rate is calculated as F(t)=i / (n+1), where F(t) is the cumulative failure rate, i is the number of failures, and n is the total number of samples.

[0011] In the step S7, it is determined whether the a values ​​in the linear equations of the acceleration temperatures are similar. If the difference between the a values ​​is ≤ 0.5, they are determined to be similar.

[0012] The specific method of step S8 is as follows: S81, substituting the cumulative failure rate = 50% into the linear equation of each accelerated temperature to obtain the logarithmic value of each median life t(0.5) under each accelerated temperature test condition; S82, calculating the antilogarithm to obtain the median life t(0.5) of the sample under each accelerated temperature test condition.

[0013] Compared with the prior art, the present invention accelerates the test time of the reliability of intermetallic compounds in wire bonding, saving time and cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 Schematic diagram of the test device in the embodiment of the present invention.

[0015] Figure 2 This is a flow chart of sample failure judgment in an embodiment of the present invention.

[0016] Figure 3 It is a test data table in the embodiment of the present invention.

[0017] Figure 4 It is a coordinate diagram showing the relationship between the accelerated failure time and the cumulative failure rate of the samples in the embodiments of the present invention in a log-normal distribution coordinate system.

[0018] Figure 5 The figure is a relationship table between the median life and the inverse of the acceleration temperature of the samples under various acceleration conditions in the embodiments of the present invention.

[0019] Figure 6 It is a coordinate diagram of the relationship between the logarithm of the median life of the samples under various acceleration conditions and the inverse of the acceleration temperature in the Arrhenius equation linear coordinate system in the embodiments of the present invention. DETAILED DESCRIPTION

[0020] The present invention will be further described below with reference to the accompanying drawings. Example

[0021] In this embodiment, gold wire bonded BGA packaging products are selected, the number of samples in each group is 10, the conventional test scheme is 150°C / 1000 hours, and 4 groups of accelerated test temperatures are selected, namely 190°C, 220°C, 240°C, and 260°C.

[0022] During the test, if Figure 1 As shown, the sample is placed on the corresponding test board, and the test board is placed in a high temperature test box, and the high temperature test box is set to the corresponding accelerated temperature for high temperature testing. At the same time, the online resistance test system tests the resistance value of the corresponding product pin in real time.

[0023] In this embodiment, the sample failure condition is set to resistance change rate (Ri-R0) / R0 ≥ 10%. Where R0 is the initial value of the resistance value, and Ri (i=1,2…) is the resistance value tested in the i-th hour. The test is carried out until all samples in each accelerated temperature group fail, during which the data acquisition system records the failure time of each sample.

[0024] The cumulative failure rate when each sample fails is obtained by the following formula: F(t) = i / (n+1), where F(t) is the cumulative failure rate, i is the number of failures, and n is the total number of samples.

[0025] Then organize the test data of temperature, product failure time, and cumulative failure rate into a table, such as Figure 3 shown.

[0026] Obtain the distribution of the failure time of each sample to the cumulative failure rate of the product, and plot the data in the test data table in the log-normal distribution coordinate system, such as Figure 4 As shown, the fitting linear equation for each acceleration temperature is obtained.

[0027] y=ax+b, where y is the cumulative failure rate, x is the logarithm of the failure time, a is the slope, and b is the intercept; Specifically: 190°C: y = 1.0175x - 8.3254 (1) 220℃: y = 1.3289x - 9.9372 (2) 240℃: y = 1.3603x - 9.5077 (3) 260℃: y = 1.0369x - 6.5968 (4) Determine whether the a values ​​in the linear equations of each accelerated temperature are similar. According to the fitting results (1), (2), (3), (4), the difference between them is ≤0.5. The slopes of the fitted equations are similar, which proves that the accelerated temperature does not change the failure mechanism of the intermetallic compound. At the same time, it can be concluded that the activation energy remains unchanged at each accelerated temperature. The similar slopes prove that the accelerated temperature does not change the failure mechanism of the intermetallic compound. At the same time, it can be concluded that the activation energy remains unchanged at each accelerated temperature, indicating that the temperature selected in the experiment effectively accelerates the failure mechanism.

[0028] Substituting the cumulative failure rate = 50% into the linear equation of each accelerated temperature, the logarithm of each median life t(0.5) under each accelerated temperature test condition is obtained. The antilogarithm can be used to obtain the median life of the sample under each accelerated temperature test condition. In this embodiment, the values ​​obtained are as follows: Figure 5 shown.

[0029] Specifically: 190℃: Ln(t(0.5))=8.67, median life t(0.5)=5847 220℃: Ln(t(0.5))=7.85, median life t(0.5)=2576 240℃: Ln(t(0.5))=7.36, median life t(0.5)=1567 260℃: Ln(t(0.5))=6.84, median life t(0.5)=938 The logarithm of the median life span, i.e. Lnt(0.5), and the inverse of the temperature value, i.e. Figure 5 The data in the Arrhenius equation are plotted in the linear coordinate system to show the relationship between the two, as shown in Figure 6 shown.

[0030] Perform a linear fit on each data point to get the fitted linear equation: y = 0.6415x - 5.1696 The slope of the straight line, 0.6415, is the activation energy value of the failure mechanism of the intermetallic compound of this sample.

[0031] The acceleration coefficient at the accelerated test temperature is obtained according to the Arrhenius equation: AF=L1 / L2=exp[Ea / k (1 / T1 - 1 / T2) ] Where AF is the acceleration factor, L1 is the life under normal conditions, L2 is the life under accelerated conditions, Ea is the activation energy value of the failure mechanism, k is the Boltzmann constant (8.62x10 -5 ), T1: conventional test temperature, T2: accelerated test temperature; The acceleration factor of 190℃ to 150℃ is: AF=exp[0.6415 / 8.62x10 -5 (1 / 150-1 / 190)]=5 The acceleration factor of 220℃ to 150℃ is: AF=exp[0.6415 / 8.62x10 -5 (1 / 150-1 / 220)]=12 The acceleration factor of 240℃ to 150℃ is: AF=exp[0.6415 / 8.62x10 -5 (1 / 150-1 / 240)]=22 The acceleration factor of 260℃ to 150℃ is: AF=exp[0.6415 / 8.62x10 -5 (1 / 150-1 / 260)]=38 The test time at the accelerated test temperature is obtained according to the acceleration coefficient: t1=t2 / AF Wherein, t1: test time at accelerated test temperature, t2: test time at normal test temperature, AF: acceleration factor.

[0032] Accelerated test time at 190°C = 1000 / 5 = 200 hours Accelerated test time at 220°C = 1000 / 12 = 83 hours Accelerated test time at 240°C = 1000 / 22 = 45 hours Accelerated test time at 260°C = 1000 / 38 = 26 hours Finally, the accelerated test solution for the intermetallic compound reliability of this BGA product is obtained: 190℃ / 200 hours or 220℃ / 83 hours or 240℃ / 45 hours or 260℃ / 26 hours.

Claims

1. An accelerated test method for reliability of intermetallic compounds in wire bonding, characterized in that: The steps include: S1, select multiple samples and multiple accelerated high temperature temperatures, and conduct high temperature tests respectively; S2, measure the resistance value of each sample pin under different high temperature environments; S3, setting the failure condition, and the time when the failure condition is reached is called the failure time; S4, calculate the cumulative failure rate of each sample when it fails; S5, arranging the test data into a table, the test data including temperature, product failure time, and cumulative failure rate; S6, obtain the distribution of the failure time of each sample to the cumulative failure rate of the product, plot the data in the test data table in a log-normal distribution coordinate system and obtain the fitted linear equation of each accelerated temperature, the fitted linear equation is y=ax+b, where y is the cumulative failure rate, x is the logarithmic value of the failure time, a is the slope, and b is the intercept; S7, perform linear fitting on the data points of each acceleration temperature to confirm whether the slopes of each straight line are similar. If so, proceed to step S8; if not, lower the selected value of the highest temperature and adjust the interval value of each temperature to repeat steps S1 to S7; S8, according to the linear equation, calculate the time when the cumulative failure rate is 50% at each accelerated temperature, that is, the median life t(0.5); S9, plot the logarithm of the median life span, i.e., Lnt(0.5), and the inverse of the temperature value in the Arrhenius equation linear coordinate system to show the relationship between the two; S10, linear fitting is performed again on each data point to obtain the slope of the straight line, which is the activation energy value of the failure mechanism of the intermetallic compound. The fitted linear equation is m=cn+d, where m is Lnt(0.5), n is (1 / T)*10000, c is the slope, that is, the activation energy value of the failure mechanism, and d is the intercept; S11, substitute the activation energy value, accelerated test temperature, and normal test temperature into the Arrhenius equation to obtain the acceleration coefficient at the accelerated test temperature, thereby obtaining the test time at the accelerated test temperature; In step S11, the method for obtaining the acceleration coefficient at the accelerated test temperature according to the Arrhenius equation is as follows: AF=L1 / L2=exp[Ea / k (1 / T1-1 / T2) ], where AF is the acceleration coefficient, L1 is the life under normal conditions, L2 is the life under accelerated conditions, Ea is the activation energy value of the failure mechanism, and k is the Boltzmann constant (8.62X10 -5 ), T1 is the conventional test temperature, T2 is the accelerated test temperature; The method for obtaining the test time at the accelerated test temperature based on the acceleration coefficient is as follows: t1=t2 / AF, where t1 is the test time at the accelerated test temperature, t2 is the test time at the conventional test temperature, and AF is the acceleration coefficient.

2. The accelerated test method for reliability of intermetallic compounds in wire bonding according to claim 1, characterized in that: In the step S1, the temperature of the accelerated high temperature is 180-300°C.

3. The accelerated test method for reliability of intermetallic compounds in wire bonding according to claim 1, characterized in that: In the step S2, the resistance value is measured once every hour.

4. The accelerated test method for reliability of intermetallic compounds in wire bonding according to claim 1, characterized in that: In step S9, the unit of the temperature value is Kelvin.

5. The accelerated test method for reliability of intermetallic compounds in wire bonding according to claim 1, characterized in that: In the step S3, the failure condition is the resistance change rate (Ri-R0) / R0 ≥ 10%; wherein R0 is the initial value of the resistance value, and Ri (i=1,2...) is the resistance value tested in the i-th hour.

6. The accelerated test method for reliability of intermetallic compounds in wire bonding according to claim 1, characterized in that: In step S4, the cumulative failure rate is calculated as F(t)=i / (n+1), where F(t) is the cumulative failure rate, i is the number of failures, and n is the total number of samples.

7. The accelerated test method for reliability of intermetallic compounds in wire bonding according to claim 1, characterized in that: In the step S7, it is determined whether the a values ​​in the linear equations of the acceleration temperatures are similar. If the difference between the a values ​​is ≤ 0.5, they are determined to be similar.

8. The accelerated test method for reliability of intermetallic compounds in wire bonding according to claim 1, characterized in that: The specific method of step S8 is as follows: S81, substituting the cumulative failure rate = 50% into the linear equation of each accelerated temperature to obtain the logarithmic value of each median life t(0.5) under each accelerated temperature test condition; S82, by calculating the antilogarithm, we can obtain the median life t(0.5) of the sample under each accelerated temperature test condition.

Images