Nanometer metal welding layer sintering quality prediction method and system based on molecular dynamics

Abstract

The invention relates to the technical field of nanometer sintered metal welding layers for semiconductor chip packaging, in particular to a nanometer metal welding layer sintering quality prediction method and system based on molecular dynamics, and the method comprises the following steps: S1, selecting materials of a chip, a nanometer metal welding layer and a substrate coating, establishing a chip-welding layer-substrate sandwich model conforming to a packaging application scene; s2, potential functions among atoms in the sandwich model are determined according to the selected materials; s3, inputting molecular modeling software and molecular dynamics parameters, and obtaining sintering result data; and S4, exporting the sintering result data, and performing visualization processing to obtain corresponding visualization information. According to the method, the visual model is established from the atomic / molecular level on the basis of the actual scene of chip packaging and the molecular dynamics method, the influence of different process parameters on the sintering quality of the nano-metal welding layer can be explored, and the method has important significance for optimizing the nano-metal sintering process parameters so as to obtain a more reliable welding layer.

Description

Technical Field

[0001] The present invention relates to the technical field of nano-sintered metal welding layers for semiconductor chip packaging, and in particular to a method and system for predicting the sintering quality of nano-metal welding layers based on molecular dynamics. Background Art

[0002] With the development of power electronic devices, the high-temperature resistance, high electrical conductivity, and high thermal conductivity of solder layers are gaining increasing attention. Nanometal sintered materials, such as nano-copper and nano-silver, are highly anticipated in the third-generation semiconductor packaging field due to their low-temperature sintering, high-temperature service, and high electrical and thermal conductivity.

[0003] CN114414631A discloses a non-destructive prediction method for the quality of weld layer sintering based on time-frequency domain analysis of electric pulse signals. The steps are: 1) pulse discharge is performed on the sintered sample; 2) the pulse signal received by the sintered sample is collected by a signal acquisition unit and transmitted to a host computer; 3) the host computer pre-processes the pulse signal; 4) the host computer performs data processing on the pre-processed pulse signal to obtain a time-domain impedance signal; 5) fast Fourier transform is performed on the time-domain impedance signal to obtain a frequency-domain impedance signal; 6) the frequency-domain impedance signal and the time-domain impedance signal are analyzed to determine whether the quality of the sintered sample is qualified. Through time-frequency domain analysis of electric pulse signals, the problem of weld layer sintering quality detection is solved, and the quality rate and service life of electronic products are improved. CN116698982A discloses a method for determining the welding quality of power semiconductor devices. By combining acoustic non-destructive scanning and cyclic thermal experiments with ultrasonic testing and highly accelerated life testing, the problem of difficult to ensure the welding quality of power semiconductor devices is solved, the detection accuracy and device reliability are improved, and refined design and cost reduction are achieved.

[0004] The sintering quality of nanometal is affected by a series of parameters such as sintering temperature, sintering pressure, and sintering time. The above methods are all tested after the power semiconductor device is welded. It is impossible to intervene in the process parameters in real time during the welding process, which may lead to batch rework or defective products flowing into subsequent links. Summary of the Invention

[0005] In view of the problem in the prior art that the welding layer is easily defective when being inspected after welding, the present invention provides a method and system for predicting the sintering quality of nano-metal welding layers based on molecular dynamics.

[0006] The present invention is achieved through the following technical solutions: A method for predicting the sintering quality of a nano-metal welding layer based on molecular dynamics comprises the following steps: S1, select the materials of chip, nano-metal solder layer and substrate coating, and establish a chip-solder layer-substrate sandwich model that meets the packaging application scenario; S2, determine the potential function between atoms in the sandwich model according to the selected material; S3, input molecular modeling software and molecular dynamics parameters to obtain sintering result data; S4, exporting the sintering result data and performing visualization processing to obtain corresponding visualization information.

[0007] Preferably, in S1, the chip material is silicon carbide, the nano-metal solder layer material is nano-copper particles, and the substrate surface plating metal is gold.

[0008] Preferably, in S2, the interactions between atoms in the coating metal and between particles in the nanometal solder layer are described by eam potential, the interactions between crystals in the chip are described by tersoff potential, and the interactions between different types of atoms are described by lj / cut potential.

[0009] Preferably, in S3, the parameters of the molecular modeling software include model size, metal particle size, boundary conditions, and the parameters of the molecular dynamics software include ensemble type, sintering temperature, sintering pressure, heating time, holding time, cooling relaxation time, shear direction and shear speed.

[0010] Preferably, the X direction is used as the shear direction, the boundary condition is set to s, and the boundary conditions in the Y direction and the Z direction are both p.

[0011] Preferably, the nvt ensemble is used to control the sintering temperature, sintering pressure, heating time, holding time, and cooling relaxation time, and the move linear instruction is used to set the chip layer shear speed.

[0012] Preferably, in S4, Ovito is used for coordinate visualization, and Excel is used to process the statistical shearing data to draw an image.

[0013] A nano-metal solder layer sintering quality prediction system based on molecular dynamics includes a model building module, a potential function building module, a sintering simulation analysis module and a visualization processing module; the model building module is used to establish a chip-solder layer-substrate sandwich model that meets the packaging application scenario, and export the model data into a data format that can be recognized by molecular dynamics simulation software; the potential function building module is used to determine the potential function between each atom in the sandwich model according to the selected material; the sintering simulation analysis module is used to input molecular modeling software and molecular dynamics parameters to obtain sintering result data; the visualization processing module is used to export the sintering result data, perform visualization processing, and obtain corresponding visualization information.

[0014] An electronic device includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method when executing the computer program.

[0015] A storage medium stores a computer program thereon, wherein the computer program implements the steps of the method when executed by a processor.

[0016] Compared with the prior art, the present invention has the following beneficial effects: The present invention provides a method for predicting the sintering quality of nano-metal solder layers based on molecular dynamics. The method is based on the actual scenario of chip packaging and applies the molecular dynamics method to establish a visualization model at the atomic / molecular level. It can explore the influence of different process parameters on the sintering quality of nano-metal solder layers, that is, changing the potential function to study different nano-metal materials and substrate coating materials, and changing various parameters to study the influence of different parameters on the sintering quality. This is of great significance for optimizing the nano-metal sintering process parameters to obtain a more reliable solder layer.

[0017] Furthermore, potential functions, or force fields, describe the interactions between particles in a system. The LAMMPS commands pair_style, pair_coeff, and pair_modify set the type and parameters of the potential function. Typically, the most time-consuming step in a molecular dynamics simulation is not the integration of force, acceleration, velocity, and displacement, but the calculation of the potential function. Different potential functions describe the interactions between atoms under different circumstances, and the correct choice of potential function is crucial to the accuracy and efficiency of the simulation in a specific scenario. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a flow chart of a method for predicting the sintering quality of a nano-metal welding layer based on molecular dynamics of the present invention; Figure 2 This is a schematic diagram of a sandwich model in a method for predicting the sintering quality of a nano-metal welding layer based on molecular dynamics according to the present invention; Figure 3 Schematic diagram of sintering and pressurization in an embodiment of the present invention; Figure 4 Schematic diagram of the shearing process in an embodiment of the present invention; Figure 5 This is a curve of the sum of the atomic stresses of the four welding layers in the X direction obtained by controlling other variables and adjusting only the sintering pressure in the embodiment of the present invention; Figure 6 1 is a schematic diagram of the sintering of the model when the sintering pressure is 15 MPa in an embodiment of the present invention; Figure 7 This is a microscopic image of atomic stress distribution when the sintering pressure is 1 MPa in an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The present invention will be further described in detail below with reference to specific embodiments, which are intended to explain the present invention rather than to limit it.

[0020] The present invention discloses a method for predicting the sintering quality of nano-metal welding layer based on molecular dynamics. Figure 1 , including the following steps: S1, select the materials of chip, nano-metal solder layer and substrate coating, and use the molecular modeling software Materials Studio to build a chip-solder layer-substrate sandwich model that meets the packaging application scenario (such as Figure 2 ), use the molecular dynamics Lammps msi command to export the model built by Materials Studio into a data file that it can recognize. The chip material is silicon carbide, the nanometal solder layer material is nanocopper particles, and the substrate surface coating metal is gold.

[0021] S2. Determine the potential functions between atoms in the sandwich model based on the selected material. The interactions between atoms in the coating metal and between particles in the nanometal solder layer are both described using the eam potential. The interactions between crystals in the chip are described using the tersoff potential. The interactions between different types of atoms are described using the lj / cut potential. Specifically, the Au.eam potential is used to describe the interactions between gold atoms in the coating metal, the Cu.eam potential is used to describe the interactions between nanocopper particles, the SiC.tersoff potential is used to describe the interactions between silicon carbide crystals in the chip, and the lj / cut potential is used to describe the interactions between different types of atoms.

[0022] In S3, input molecular modeling software and molecular dynamics parameters to obtain sintering result data. Specifically, set the X-direction as the shear direction, the boundary condition to s, and the boundary conditions in the Y and Z directions to p. Write the lammps.in file code to set the sintering pressure, sintering temperature, sintering time, and shearing method. Use the nvt ensemble to control the sintering temperature, sintering pressure, heating time, holding time, and cooling relaxation time. Use the move linear command to set the chip layer shear velocity.

[0023] S4, the sintering result data was exported and coordinate visualization was performed using Ovito, and the statistical shear data was processed using Excel to draw images and obtain the corresponding visualization information.

[0024] Example In an example of shear strength testing of a sandwich model of a silicon carbide chip, a nano-sintered copper solder layer, and a gold-plated substrate at different sintering pressures, the molecular dynamics-based method for simulating the reliability of the nano-metal solder layer is as follows: Step 1: Use Materials Studio to build a chip-solder layer-substrate model. Import the chip material into the SiC unit cell model that comes with the software, and use the supercell command to generate a rectangular supercell with a length of 6nm, a width of 6nm, and a height of 2nm as the upper chip layer. Then import the Au unit cell model that comes with the software, and use the supercell command to generate a rectangular supercell with a length of 6nm, a width of 6nm, and a height of 3nm as the bottom coating substrate. Import the Cu unit cell model that comes with the software, and use the Nanocluster command to generate 9 nano-copper balls with a particle size of 2nm, and arrange them evenly above the substrate. Use the layer command to generate a rectangular simulation box with a length of 6.1175nm, a width of 6.1175nm, and a height of 7.34193nm for the chip layer, nano-copper balls, and substrate layer; Step 2: Use Lammps's msi command to convert the model built in step 1 into a data model that Lammps can recognize; Step 3: Write the in file, set the boundary condition in the x direction (shear direction) to s, the boundary conditions in the y and z directions to p, and import the data model file; Step 4: The interaction between the gold atoms in the coating is described by the eam potential, and the built-in Au_u3.eam file is imported. The interaction between the nano-copper atoms is also described by the eam potential, and the built-in Cu_u3.eam file is imported. The interaction between the SiC crystals is described by the tersoff potential, and the built-in SiC.tersoff file is imported. The interaction between different types of atoms is described by the lj / cut potential. The gold atoms have no direct contact with the chip and are far away. For the sake of simulation efficiency, no interaction is set. The parameters between Cu and C atoms are set to 0.0432 and 3, the parameters between Cu and Si atoms are set to 0.0845 and 3.16, and the parameters between Au and Cu atoms are set to 0.425 and 2.48. Step 5: Use the stress / atom command to calculate the stress of each atom in the nano-copper layer, and use the dump command to output the coordinate information every 100 steps; Step 6: Use the NVT ensemble to control the system temperature to rise from 300K to 500K, with a heating time of 1000fs; Step 7: Use the nvt ensemble to maintain the temperature at 500K, and use the addforce command to apply a pressure value of 0.0225 (equivalent to 1 MPa), and maintain the sintering time at 2000 fs; Step 8: Use the NVT ensemble to control the system temperature from 500K to 300K, with a cooling time of 1000fs; Step 9: Use the nvt ensemble to maintain the temperature at 300K and the cooling relaxation time at 4000fs; Step 10: Use the nvt ensemble to maintain the temperature of the nanocopper layer at 300K. Use the move linear command to set the chip layer shear velocity to 1 angstrom / picosecond in the positive x direction and maintain it at 4500 fs. Export the shear stress data. Step 11: Use Ovito to visualize the coordinates and Excel to process the statistical shear data and draw the image; Step 12: Change the pressure values ​​in step 7 to 5MPa, 10MPa, and 15MPa, and repeat steps 8 to 11. The test results of the sum of atomic stresses of the weld layer in the X direction are as follows: Figure 5 shown.

[0025] Figure 3 as well as Figure 4 As shown in the figure, the sintering and shearing processes are simulated on a micro scale, which is convenient for direct observation and research; a series of factors affecting the sintering quality of the solder layer, such as sintering temperature, temperature gradient, sintering time, sintering pressure, nanometal size, substrate material, chip material, etc., are set as variables, and the model is highly usable.

[0026] Figure 5 The sum curve of the atomic stress in the X direction of the four weld layers obtained by controlling other variables and adjusting only the sintering pressure is shown. Figure 6 The reason why the shear strength decreases when the sintering pressure is too high is revealed from a microscopic level: excessively high pressure will cause the silicon carbide chip to deform, resulting in uneven pressure application, and the chip and the solder layer will no longer fit tightly, so the strength will decrease.

[0027] Figure 7 It shows the microscopic image of atomic stress distribution observed by cross-section processing in Ovito post-processing.

[0028] The present invention also discloses a nano-metal welding layer sintering quality prediction system based on molecular dynamics, which includes a model building module, a potential function building module, a sintering simulation analysis module and a visualization processing module; the model building module is used to establish a chip-welding layer-substrate sandwich model that meets the packaging application scenario, and export the model data into a data format that can be recognized by molecular dynamics simulation software; the potential function building module is used to determine the potential function between each atom in the sandwich model according to the selected material; the sintering simulation analysis module is used to input molecular modeling software and molecular dynamics parameters to obtain sintering result data; the visualization processing module is used to export the sintering result data, perform visualization processing, and obtain corresponding visualization information.

[0029] The present invention also discloses an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method when executing the computer program.

[0030] The present invention also discloses a storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the method are implemented.

[0031] The above description is merely a preferred embodiment of the present invention and is not intended to impose any limitation on the technical solution of the present invention. Those skilled in the art should understand that, without departing from the spirit and principles of the present invention, the technical solution can also be subjected to several simple modifications and replacements, and these modifications and replacements are also within the scope of protection covered by the claims.

Claims

1. A method for predicting the sintering quality of nano-metal welding layers based on molecular dynamics, characterized in that: The following steps are involved: S1, select the materials of chip, nano-metal solder layer and substrate coating, and establish a chip-solder layer-substrate sandwich model that meets the packaging application scenario; S2, determine the potential function between atoms in the sandwich model according to the selected material; S3, input molecular modeling software and molecular dynamics parameters to obtain sintering result data; S4, exporting the sintering result data and performing visualization processing to obtain corresponding visualization information.

2. The method for predicting the sintering quality of nano-metal welding layer based on molecular dynamics according to claim 1, characterized in that: In S1, the chip material is silicon carbide, the nano-metal solder layer material is nano-copper particles, and the substrate surface plating metal is gold.

3. The method for predicting the sintering quality of nano-metal welding layers based on molecular dynamics according to claim 2, characterized in that: In S2, the interactions between atoms in the coating metal and the interactions between particles in the nanometal solder layer are described by the eam potential, the interactions between crystals in the chip are described by the tersoff potential, and the interactions between different types of atoms are described by the lj / cut potential.

4. The method for predicting the sintering quality of nano-metal welding layer based on molecular dynamics according to claim 1, characterized in that: In S3, the parameters of the molecular modeling software include model size, metal particle size, and boundary conditions; the parameters of the molecular dynamics software include ensemble type, sintering temperature, sintering pressure, heating time, holding time, cooling relaxation time, shear direction, and shear speed.

5. The method for predicting the sintering quality of nano-metal welding layer based on molecular dynamics according to claim 4, characterized in that: The X direction is set as the shear direction, the boundary condition is set to s, and the boundary conditions in the Y and Z directions are both p.

6. The method for predicting the sintering quality of nano-metal welding layers based on molecular dynamics according to claim 4, wherein: The nvt ensemble is used to control the sintering temperature, sintering pressure, heating time, holding time, and cooling relaxation time, and the movelinear instruction is used to set the chip layer shear speed.

7. The method for predicting the sintering quality of nano-metal welding layer based on molecular dynamics according to claim 1, characterized in that: In S4, Ovito was used for coordinate visualization, and Excel was used to process the statistical shear data and draw the images.

8. A nano-metal welding layer sintering quality prediction system based on molecular dynamics, characterized in that: It includes a model building module, a potential function building module, a sintering simulation analysis module, and a visualization processing module. The model building module is used to establish a chip-solder layer-substrate sandwich model that meets the packaging application scenario and export the model data into a data format that can be recognized by molecular dynamics simulation software. The potential function establishment module is used to determine the potential function between each atom in the sandwich model according to the selected material; The sintering simulation analysis module is used to input molecular modeling software and molecular dynamics parameters to obtain sintering result data; The visualization processing module is used to export the sintering result data and perform visualization processing to obtain the corresponding visualization information.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images