High-performance alloy bump material and IC packaging process

Abstract

The invention relates to a high-performance alloy bump material and an IC packaging technology, and the technology comprises the following steps: employing a metallographic microscope to detect the internal crystal grain structure of an alloy bump, and dividing the alloy bump into at least eight heating regions; according to detection and region division results, working parameters of the electromagnetic induction heating device are dynamically adjusted, and accurate control over the alloy bump heating process is achieved; and a metallographic microscope and a four-probe tester are used for detecting the structure and resistivity of the heated crystal grains, so that the product quality is ensured. The process has the characteristics of subdivision of a heating area, dynamic adjustment of parameters, realization of uniform heating, improvement of heating efficiency, energy conservation and emission reduction; and meanwhile, the product reliability is improved.

Description

Technical Field

[0001] The invention relates to a high-performance alloy bump material and an IC packaging process. Background Art

[0002] In traditional IC packaging processes, alloy bumps are usually processed by uniform heating; however, this method has certain shortcomings, such as uneven heating and difficulty in controlling the grain structure. The uneven grain structure inside the alloy bump will seriously affect the performance of the bump after packaging.

[0003] Therefore, a high-performance alloy bump material and IC packaging process are needed to solve the problems of uneven heating and difficult to control grain structure. Summary of the invention

[0004] The purpose of the present invention is to solve the above deficiencies in the prior art, provide a high-performance alloy bump material and an IC packaging process, and achieve precise control of the internal grain structure of the alloy bump through a special IC packaging process combined with the selection of alloy bump and electroplating materials. The following is a specific solution:

[0005] An IC packaging process comprises the following methods: detecting the internal grain structure of an alloy bump before heating treatment by means of a metallographic microscope, and dividing the alloy bump into at least eight heating areas; dynamically adjusting the working parameters of an electromagnetic induction heating device according to the heating areas of the alloy bump; recording the working parameters into the electromagnetic induction heating device; heating the alloy bump by means of the electromagnetic induction heating device according to the working parameters, and continuously detecting the alloy bump by means of a metallographic microscope during the heating process; dynamically adjusting the working parameters of the magnetic induction heating device according to the detection result of the metallographic microscope; detecting the internal grain structure of the alloy bump after heating treatment by means of a metallographic microscope; performing resistivity test on the alloy bump by means of a four-probe tester; and electroplating the tested alloy bump.

[0006] The method of detecting the internal grain structure of the alloy bump before the heat treatment by metallographic microscope and dividing the alloy bump into at least eight heating areas includes: classifying the detection results of the metallographic microscope by a periodic action algorithm, and dividing the alloy bump into at least eight heating areas according to the classification. This algorithm can more accurately identify the microstructural features inside the alloy bump by iteratively optimizing the cluster center, thereby achieving more precise temperature control, ensuring the uniformity and accuracy of the heating treatment, and not only improving the heating efficiency, but also improving the quality and stability of the product.

[0007] The periodic action algorithm is specifically as follows: randomly select 8 data points as initial cluster centers; assign each data point to the nearest cluster center to form 8 clusters; recalculate the center of each cluster; repeat the above steps until the cluster center no longer changes or reaches a predetermined number of iterations, wherein the assignment of each data point to the nearest cluster center to form 8 clusters is calculated based on the following formula:

[0008] C j ={x i :||x i -c j ||≤||x i -c k ||for all k≠j}, where C j For clusters,

[0009] x i is the data point, c j is the cluster center; the recalculation of the center of each cluster is based on the following formula: |C j | is the number of data points in the cluster. The periodic action algorithm is used to classify the inspection results of the metallographic microscope, which is faster and more accurate.

[0010] The method of dynamically adjusting the working parameters of the electromagnetic induction heating device according to the heating area of ​​the alloy bump includes: automatically identifying the heating area through a deep learning algorithm, and dynamically adjusting the electromagnetic field intensity and heating time of the electromagnetic induction heating device according to the heating area, and also includes the following steps: using a deep learning model to extract features from image data to help the model identify and classify different heating areas; using a labeled data set to train the deep learning model, during the training process, the model will learn how to predict the heating area according to the input image data, and output the corresponding electromagnetic field intensity and heating time; the model will predict the heating area according to the real-time collected image data, and automatically adjust the parameters of the electromagnetic induction heating device according to the prediction results; the algorithms of the above steps are all based on the following formula:

[0011] Among them, y i is the true value, is the model prediction value, N is the number of samples, and through the application of deep learning algorithm, real-time optimization of the heating process is achieved.

[0012] The alloy bump is made of a high-lead alloy, which has higher welding performance and reliability.

[0013] The material of the electroplating layer of the alloy bump is one of gold-tin alloy, nickel-gold alloy or copper-nickel-gold alloy. These alloys have excellent electrical conductivity and welding performance and are convenient for electroplating processing.

[0014] Beneficial effects: The present invention realizes precise control of the internal grain structure of the alloy bump through a special IC packaging process, detects during heating through a metallographic microscope, and divides it into 8 areas. The detection results are fed back to the electromagnetic induction heating device to ensure the uniformity of the heating treatment and make the grain structure more uniform. This optimization not only improves the heating efficiency and makes the heating more uniform, but also has the effect of energy saving and emission reduction;

[0015] By testing the grain structure and resistivity using a metallographic microscope and a four-probe tester, potential quality problems can be discovered in a timely manner, thereby improving product reliability.

[0016] In general, after precisely controlled heating and electroplating processes, the resistivity of the alloy bumps is effectively reduced, and IC packaging is faster. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a flow chart of high-performance alloy bump materials and IC packaging processes;

[0018] Figure 2 It is a schematic diagram of the device structure of high-performance alloy bump material and IC packaging process. DETAILED DESCRIPTION

[0019] In order to deepen the understanding of the present invention, the present invention will be further described in detail below in conjunction with embodiments and drawings. The embodiments are only used to explain the present invention and do not constitute a limitation on the protection scope of the present invention.

[0020] In this embodiment:

[0021] Please see Figure 1-2 , an IC packaging process, comprising the following methods: detecting the internal grain structure of an alloy bump before heating treatment through a metallographic microscope, and dividing the alloy bump into at least eight heating areas; dynamically adjusting the working parameters of an electromagnetic induction heating device according to the heating areas of the alloy bump; entering the working parameters into the electromagnetic induction heating device; heating the alloy bump using the electromagnetic induction heating device according to the working parameters, and continuously detecting the alloy bump through a metallographic microscope during the heating process; dynamically adjusting the working parameters of the magnetic induction heating device according to the detection results of the metallographic microscope; detecting the internal grain structure of the alloy bump after heating treatment through a metallographic microscope; performing resistivity test on the alloy bump using a four-probe tester; and electroplating the tested alloy bump. In this embodiment, eight equally divided heating areas are used, wherein the areas of the eight heating areas are equal and the whole is in the shape of eight equal parts.

[0022] The internal grain structure of the alloy bump before the heat treatment is detected by a metallographic microscope, and the alloy bump is divided into at least eight heating areas, including: the detection results of the metallographic microscope are classified by a periodic action algorithm, and the alloy bump is divided into at least eight heating areas according to the classification.

[0023] The specific periodic action algorithm is as follows: randomly select 8 data points as the initial cluster centers; assign each data point to the nearest cluster center to form 8 clusters; recalculate the center of each cluster; repeat the above steps until the cluster center no longer changes or reaches a predetermined number of iterations, where each data point is assigned to the nearest cluster center to form 8 clusters based on the following formula:

[0024] C j ={x i :||x i -c j ||≤||x i -c k ||for all k≠j}, where C j is a cluster, x i is the data point, c j is the cluster center; recalculate the center of each cluster based on the following formula: |C j | is the number of data points in the cluster.

[0025] The method of dynamically adjusting the working parameters of the electromagnetic induction heating device according to the heating area of ​​the alloy bump includes: automatically identifying the heating area through a deep learning algorithm, and dynamically adjusting the electromagnetic field intensity and heating time of the electromagnetic induction heating device according to the heating area, and also includes the following steps: using a deep learning model to extract features from image data to help the model identify and classify different heating areas; using a labeled data set to train the deep learning model, during the training process, the model will learn how to predict the heating area based on the input image data and output the corresponding electromagnetic field intensity and heating time; the model will predict the heating area based on the real-time collected image data, and automatically adjust the parameters of the electromagnetic induction heating device according to the prediction results; the algorithms of the above steps are all based on the following formula:

[0026] Among them, y i is the true value, is the model prediction value, and N is the number of samples.

[0027] The material of the alloy bump is a high-lead alloy, and the material of the electroplating layer of the alloy bump is one of a gold-tin alloy, a nickel-gold alloy, or a copper-nickel-gold alloy. In this embodiment, a copper-nickel-gold alloy is used as the material of the electroplating layer. The electroplating process is as follows: clean the alloy bump; put the cleaned alloy bump into an electroplating solution containing a copper-nickel-gold alloy; turn on the power for electroplating, and the current density is controlled at 2.5A / dm 2 The electroplating time is 1-2 minutes; after the electroplating is completed, the alloy bumps are taken out, cleaned and dried.

[0028] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

Claims

1. An IC packaging process, characterized in that: The invention comprises the following methods: using a metallographic microscope to detect the internal grain structure of the alloy bump before the heating treatment, and dividing the alloy bump into at least eight heating areas; dynamically adjusting the working parameters of an electromagnetic induction heating device according to the heating areas of the alloy bump; recording the working parameters into the electromagnetic induction heating device; using the electromagnetic induction heating device to heat the alloy bump according to the working parameters, and continuously detecting the alloy bump through a metallographic microscope during the heating process; dynamically adjusting the working parameters of the magnetic induction heating device according to the detection results of the metallographic microscope; detecting the internal grain structure of the alloy bump after the heating treatment through a metallographic microscope; using a four-probe tester to perform a resistivity test on the alloy bump; and electroplating the tested alloy bump.

2. An IC packaging process according to claim 1, characterized in that: The detecting of the internal grain structure of the alloy bump before the heat treatment by metallographic microscope and dividing the alloy bump into at least eight heating areas comprises: classifying the detection result of the metallographic microscope by periodic action algorithm and dividing the alloy bump into at least eight heating areas according to the classification.

3. An IC packaging process according to claim 2, characterized in that: The periodic action algorithm is specifically as follows: randomly select 8 data points as initial cluster centers; assign each data point to the nearest cluster center to form 8 clusters; recalculate the center of each cluster; repeat the above steps until the cluster center no longer changes or reaches a predetermined number of iterations, wherein the assignment of each data point to the nearest cluster center to form 8 clusters is calculated based on the following formula: C j ={x i :||x i -c j ||≤||x i -c k ||for all k≠j}, where C j is a cluster, x i is the data point, c j is the cluster center; the recalculation of the center of each cluster is based on the following formula: |C j | is the number of data points in the cluster.

4. The IC packaging process according to claim 1, characterized in that: The method of dynamically adjusting the working parameters of the electromagnetic induction heating device according to the heating area of ​​the alloy bump includes: automatically identifying the heating area through a deep learning algorithm, and dynamically adjusting the electromagnetic field intensity and heating time of the electromagnetic induction heating device according to the heating area, and also includes the following steps: using a deep learning model to extract features from image data to help the model identify and classify different heating areas; using a labeled data set to train the deep learning model, during the training process, the model will learn how to predict the heating area according to the input image data, and output the corresponding electromagnetic field intensity and heating time; the model will predict the heating area according to the real-time collected image data, and automatically adjust the parameters of the electromagnetic induction heating device according to the prediction results; the algorithms of the above steps are all based on the following formula: Among them, y i is the true value, is the model prediction value, and N is the number of samples.

5. A high-performance alloy bump material, used to implement the IC packaging process of any one of claims 1 to 4, characterized in that: The material of the alloy bump is a high-lead alloy.

6. The high performance alloy bump material according to claim 5, characterized in that: The material of the electroplating layer of the alloy bump is one of gold-tin alloy, nickel-gold alloy or copper-nickel-gold alloy.

Images