Methods for predicting crack initiation rate, methods for selecting resin materials, and methods for manufacturing electronic components.
By measuring the fracture toughness and thermal expansion rate of resin materials, combined with simulation and tensile tests, suitable resin materials can be predicted and selected, solving the problem of high accuracy in the crack generation rate of resin layers in electronic components, and manufacturing parts with reduced cracks.
Patent Information
- Application Number
- CN202480091131.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-14
- Publication Date
- 2026-11-17
AI Technical Summary
Existing technologies struggle to accurately predict the crack initiation rate in resin layers adjacent to metal wiring in electronic components, and it is also difficult to select resin materials that effectively suppress crack formation.
By measuring the fracture toughness and thermal expansion rate of resin materials, calculating the stress ratio using simulation tests, and combining tensile tests and reliability tests, the crack initiation rate can be predicted and suitable resin materials can be selected.
It achieves high-precision prediction of crack initiation rate in resin layer and selects resin materials that can effectively suppress cracks, thus manufacturing electronic components with reduced crack initiation.
Smart Images

Figure CN123126100A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a method for predicting the crack initiation rate, a method for selecting resin materials, and a method for manufacturing electronic components. Background Technology
[0002] In electronic components such as semiconductor packages where semiconductor elements are mounted on printed circuit boards (PCBs) via solder, a resin layer is sometimes formed to prevent solder from adhering to unwanted portions of the metal wiring on the PCB (Patent Document 1). In electronic components, the resin material used to form the resin layer adjacent to the metal wiring is preferably selected to suppress the crack initiation rate in the resin layer. For example, Patent Document 2 discloses a method for predicting the crack initiation rate in a resin layer adjacent to metal wiring.
[0003] Previous technical documents Patent documents Patent Document 1: Japanese Patent Application Publication No. 2021-005586 Patent Document 2: Japanese Patent Application Publication No. 2020-188243 Summary of the Invention
[0004] The technical problem to be solved by the invention This invention relates to a method for predicting the crack initiation rate in a resin layer adjacent to metal wiring in electronic components with higher accuracy. The invention also relates to a method for selecting resin materials with high accuracy that can suppress the initiation of cracks in a resin layer adjacent to metal wiring in electronic components.
[0005] means for solving technical problems This invention includes the following: [1] A method for predicting the initiation rate of cracks in a resin layer formed of resin material adjacent to metal wiring in an electronic component, comprising the following steps: The crack initiation rate is predicted based on the fracture toughness value of the resin material. [2] A method for predicting the initiation rate of cracks in a resin layer formed of resin material adjacent to metal wiring in an electronic component, comprising the following steps: The crack initiation rate is predicted based on the ratio S2 / S1 of the fracture toughness value S2 of the resin material to the stress S1 applied to the resin layer in the electronic component. [3] According to the method described in [2], it further includes the following steps: The stress S1 is calculated through simulation experiments based on the thermal expansion coefficient of the resin material. [4] A method for selecting a resin material in an electronic component having metal wiring and a resin layer formed of resin material disposed adjacent to the metal wiring, the method comprising the following steps: The resin material is selected based on its fracture toughness value. [5] A method for selecting a resin material in an electronic component having metal wiring and a resin layer formed of resin material adjacent to the metal wiring, comprising the following steps: The resin material is selected based on the ratio S2 / S1 of the fracture toughness value S2 of the resin material to the stress S1 applied to the resin layer in the electronic component. [6] According to the method described in [5], it further includes the following steps: The stress S1 is calculated through simulation experiments based on the thermal expansion coefficient of the resin material. [7] A method for manufacturing an electronic component having metal wiring and a resin layer formed of resin material adjacent to the metal wiring, comprising the following steps: The resin material is selected by any one of the methods in [4] to [6]; and The resin layer, formed from the selected resin material, is disposed adjacent to the metal wiring.
[0013] Invention Effects It can predict the crack initiation rate in resin layers adjacent to metal wiring in electronic components with higher accuracy. Furthermore, it can select resin materials that can suppress the formation of cracks in resin layers adjacent to metal wiring in electronic components with high accuracy. Attached Figure Description
[0014] Figure 1 This is a schematic diagram illustrating an example of a tensile test method used to assess fracture toughness.
[0015] Figure 2 This is a schematic diagram illustrating an example of the front part of a notch.
[0016] Figure 3 This is a schematic cross-sectional view of the analytical model of the electronic component used in the simulation test to calculate stress S1. Detailed Implementation
[0017] This invention is not limited to the following examples.
[0018] One example of a method for predicting the crack initiation rate in a resin layer formed of resin material and disposed adjacent to metal wiring in an electronic component includes the step of predicting the crack initiation rate based on the fracture toughness value S2 of the resin material. Another example of a method for predicting the crack initiation rate in a resin layer includes the step of predicting the crack initiation rate based on the ratio S2 / S1 of the fracture toughness value S2 of the resin material to the stress S1 applied to the resin layer in the electronic component. The crack initiation rate in the resin layer shows a relatively high correlation with the fracture toughness value of the resin material and the ratio S2 / S1. Therefore, for example, based on a regression line representing the relationship between the crack initiation rate and the fracture toughness value or the ratio S2 / S1 of the resin material, the crack initiation rate can be predicted with high accuracy from the fracture toughness value S2 or the ratio S2 / S1 of the resin material.
[0019] Figure 1 This is a schematic diagram illustrating an example of a tensile testing method used to evaluate the fracture toughness of resin materials. For example... Figure 1 As shown, the strip-shaped test piece 50 is held at both ends along its length by a pair of clamps 41 and 42. The test piece 50 has, for example, a width B of 9–11 mm and a length of 25–35 mm. The test piece 50 may have a thickness t of 0.01–0.2 mm. The distance Y between the clamps 41 and 42 is, for example, 10–30 mm.
[0020] Two notches 5N are formed at the center of the test piece 50 along its length, extending from both ends of the test piece 50 along a direction perpendicular to the length of the test piece 50. The ligament L, which is the distance between the front ends of the two notches 5N, can be B / 3 or more and 5t or less, or B / 3 or more and 3t or less.
[0021] The test piece 50 with notches 5N is prepared, for example, by a method including the following steps: forming a film-shaped molded body made of the same resin material as the resin material forming the resin layer; cutting strip-shaped test pieces 50 from the film-shaped molded body; and forming two notches 5N from both ends of the test piece 50 at the central position in the length direction of the test piece 50, which are perpendicular to the length direction of the test piece 50.
[0022] The 5N notch can be formed by laser irradiation. Typically, the resin material that forms the resin layer of electronic components such as semiconductor packages is a brittle material, but even for test pieces made of brittle materials, it is easy to form small notches with few cracks using a laser. Figure 2This is a schematic diagram showing the front end of the notch 5N. The radius r of the front end of the notch 5N can be, for example, less than 50 μm, less than 30 μm, or less than 10 μm. The maximum value of the width of the notch 5N (the width in the direction perpendicular to the length of the notch) can be, for example, less than 100 μm, less than 50 μm, or more than 1 μm. The size of the radius r of the front end of the notch 5N and the size of the width of the notch 5N can be adjusted according to the size of the laser output, etc.
[0023] The tensile test is conducted at 25°C and 50% relative humidity, starting from the tip of the 5N notch and continuing until the specimen breaks. The tensile speed can be, for example, 0.1 mm / min or more and 100 mm / min or 0.5 mm / min.
[0024] Based on the load-displacement curve obtained from the tensile test results, the fracture energy U accompanying the fracture of the specimen is calculated. The area between the load-displacement curve and the displacement axis corresponds to the fracture energy U. Based on the fracture energy U, the ligament length L, and the specimen thickness t, the following formula is used: S2=U / Lt Calculate the fracture toughness value S2.
[0025] Here, the fracture energy U can typically be decomposed into a fraction containing the plastic deformation energy w, as shown in the following equation. p The term includes the energy w required for crack propagation from the notch. e Items.
[0026] U=w p βL 2 t+w e Lt Brittle materials, such as resin materials that form resin layers in electronic components, do not undergo substantial plastic deformation during crack propagation, therefore the deformation rate (w) can be ignored. p The terms. Therefore, U / Lt and w e They are essentially equal, by U / Lt (≈w e The value of ) is regarded as the fracture toughness value, which can be used to evaluate the fracture toughness of resin materials.
[0027] Tensile tests can be performed on multiple specimens with varying distances (ligament L) between the two notched 5N ends. Based on the load-displacement curves obtained from these specimens with different ligament L values, the fracture energy U and the fracture energy per unit area U / Lt can be calculated, allowing for a more accurate assessment of the resin material's fracture toughness. The average fracture energy per unit area U / Lt from the multiple specimens with different ligament L values can be considered the fracture toughness value of the resin material.
[0028] The stress S1 applied to the resin layer in an electronic component can be calculated, for example, based on the thermal expansion coefficient of the resin material through simulation experiments. The stress S1 can be the maximum stress within the resin layer at a specified temperature (e.g., -65°C) calculated through simulation experiments. The simulation experiment can be a three-dimensional nonlinear structural analysis. The simulation experiment can be performed on an analysis model of an electronic component with the same structure as the electronic component whose crack initiation rate is to be predicted. Alternatively, the simulation experiment can be performed on an analysis model of an electronic component with a simplified arbitrary structure. The analysis model can be a quarter-scale model. An example of software for nonlinear structural analysis is MSC.Marc2023.4. Examples of conditions that can be set in the simulation experiment include the thermal expansion coefficient of the resin material, the elastic modulus, the stress-free temperature condition, and boundary conditions. The stress-free temperature condition can be, for example, the curing temperature of the underfill adhesive.
[0029] The thermal expansion coefficient (ppm / K) of the resin material can be measured, for example, using a thermal stress-strain measuring device (TMA, such as Seiko Instruments Inc., model: TMA / SS6100). For example, the thermal expansion coefficient within the range of -65°C to 150°C, calculated from the dimensional changes of a test piece heated from -65°C to 300°C under a load of 20 mN and a heating rate of 10°C / min, can be used in simulation tests. The elastic modulus (GPa) of the resin composition can be measured, for example, using a viscoelasticity measuring device (DMA, TA Instruments, model: RSA-G2). For example, the test mode can be set to tensile, the test temperature can be set to -65°C to 300°C, the heating rate can be set to 5°C / min, the clamp spacing can be set to 15 mm, the test frequency can be set to 1 Hz, and the test piece size can be a strip shape with a width of 10 mm and a thickness of approximately 18 μm. Simulation tests are possible. For example, the elastic modulus within the range of -65°C to 200°C can be used in simulation tests.
[0030] If the fracture toughness value S2 of the resin material is greater than or equal to the value of S2 / S1, the crack initiation rate in the resin layer tends to decrease. Therefore, it can be predicted that when the fracture toughness value S2 and / or greater than the value of S2 / S1 increases, the crack initiation rate in the resin layer will relatively decrease.
[0031] Methods for predicting crack initiation rates include, for example, the following steps: determining the fracture toughness value S2 of multiple (more than 4 and less than 20) resin materials; determining the crack initiation rate C in the resin layer of an electronic component having a resin layer formed from these resin materials during a reliability test involving repeated thermal stress; and determining the regression line representing the relationship between the crack initiation rate C and the fracture toughness value S2 using the least squares method. C (%) = α1 × S2 + β1 The constants α1 and β1 are given in the formula; and the crack initiation rate C is predicted from the fracture toughness value S2 according to this formula. Usually, α1 is negative and β1 is positive.
[0032] The method for predicting crack initiation rate includes the following steps: determining the fracture toughness value S2 of multiple (e.g., more than 4 and less than 20) resin materials; determining the stress S1 applied to the resin layer through simulation tests associated with an analytical model of an electronic component having a resin layer formed from these resin materials; determining the crack initiation rate C in the resin layer when the electronic component having a resin layer formed from these resin materials is subjected to a reliability test with repeated thermal processes; and determining the regression line representing the relationship between the crack initiation rate C and the fracture toughness value S2 using the least squares method. C (%) = α2 × (S2 / S1) + β2 The constants α2 and β2 are given in the formula; and the crack initiation rate C is predicted from the fracture toughness value S2 and stress S1 according to this formula. Usually, α2 is negative and β2 is positive.
[0033] Here, the crack initiation rate C is defined as the proportion of crack-initiating sites in multiple locations of the resin layer, as shown in the following formula.
[0034] C (%) = (n1 / n) o )×100 In the formula, n0 represents the number of observation locations, and n1 represents the number of locations where cracks were confirmed to have occurred. For example, in the case of a resin layer (solder resist, etc.) adjacent to multiple solder balls, one location of the resin layer can be observed for each solder ball.
[0035] The resin material for forming the resin layer can also be selected based on the fracture toughness value S2 or the ratio S2 / S1. The method for selecting the resin material may include, for example, the following steps: preparing a variety of candidate resin materials; determining the fracture toughness value S2 or the ratio S2 / S1 for these candidate resin materials; and selecting the resin material with the largest fracture toughness value S2 or the ratio S2 / S1. Alternatively, the resin material can be selected based on the fracture toughness value S2 or the ratio S2 / S1, such that the crack initiation rate predicted by the above method is any target value (e.g., 10%, 5%, 3%, 2%, or 1%) or less.
[0036] The resin material can be a cured product of a curable resin composition (photosensitive resin composition or thermosetting resin composition). The photosensitive resin composition may, for example, contain one or more resins selected from the group consisting of epoxy resin, polybenzoxazole (PBO) resin, phenolic varnish resin, polyimide resin and benzocyclobutene resin, and a photoacid generator (also known as a "photocationic polymerization initiator").
[0037] Resin materials or curable resin compositions may contain fillers. The fracture toughness of resin materials containing fillers can also be evaluated with high accuracy using the methods involved in this invention. Examples of fillers include silica, alumina, boron nitride, and aluminum nitride.
[0038] By employing a method that includes the step of placing a resin layer formed from a selected resin material adjacent to metal wiring, electronic components in which the generation of cracks in the resin layer is suppressed can be easily manufactured. When the resin material is a cured product of a photosensitive resin composition, the resin layer is formed, for example, by a method including the steps of: forming a film of the photosensitive resin composition; and forming a resin layer with a pattern containing openings by exposing the film and subsequently developing it. The developed resin layer can be heated as needed.
[0039] <Verification Experiment> The following describes the validation experiments related to the method for predicting crack initiation rate.
[0040] 1. Evaluation of photosensitive resin compositions Film-like photosensitive resin compositions 1, 2, 3, and 4 were prepared. The stress S1, fracture toughness value S2, and breaking stress S3 of the resin layers formed from each photosensitive resin composition were determined using the following method.
[0041] (1) Stress S1 A 18 μm thick cured film was formed by exposure and thermal curing of various photosensitive resin compositions. Using a thermal stress-strain measuring apparatus (TMA, Seiko Instruments Inc., model: TMA / SS6100), under a load of 20 mN and a heating rate of 10 °C / min, the dimensional changes of test pieces cut from the cured film were measured within the range of -65 °C to 300 °C. Based on the measurement results, the coefficient of thermal expansion (ppm / K) within the range of -65 °C to 150 °C was determined.
[0042] Through simulation experiments using the measured values of thermal expansion, the thermal expansion coefficient was determined to be... Figure 3 The stress S1 is applied to the resin layer at -65°C in the analytical model of the electronic component shown. Figure 3The analysis model 100 shown comprises: a circuit board 1 having a substrate 10 and a plurality of metal wirings 15 (copper pads) arranged on the substrate 10; a chip 3 mounted on the circuit board 1; a resin layer 5, a solder mask layer disposed adjacent to the metal wirings 15 (copper pads) on the circuit board 1 and having circular openings exposing the metal wirings 15; solder ball portions 20, filling the openings of the resin layer 5 and connecting the metal wirings 15 and the chip 3; and a bottom filling portion 7, filling the space between the resin layer 5 and the chip 3. The substrate 10 has a main surface with a length of 64 mm and a width of 64 mm. The chip 3 has a main surface with a length of 20 mm and a width of 20 mm. The metal wirings 15 have a circular main surface with a diameter of 80 μm and a thickness of 18 μm. A plurality of metal wirings 15 are arranged on the substrate 10, with an interval of 0.1 mm between adjacent metal wirings 15. The metal wirings 15 are arranged in a grid pattern of 200 vertically and 200 horizontally. The resin layer 5 (solder resist layer) has a thickness of 36 μm. The wall surface of the opening in the resin layer 5 is inclined relative to the thickness direction of the resin layer 5. The diameter of the opening is 40 μm at the position closest to the metal wiring 15 and 50 μm at the position closest to the bottom filler 7. The bottom filler 7 has a rounded corner 7a covering the side of the chip 3. The thickness of the bottom filler 7 between the chip 3 and the resin layer 5 is 50 μm. Regarding the resin composition, in order to apply the viscoelastic properties, a viscoelastic measuring device is used in the same way as the above-described method for measuring the elastic modulus. The master curve is synthesized by combining the results measured in tensile mode at frequencies of 1, 2, 4, 10, and 20 Hz. After approximation using the WLF rule of the time-temperature conversion rule, it is converted into a 10-term Prony series and input into Marc.
[0043] A three-dimensional simulation experiment was performed using the nonlinear structural analysis software MSC.Marc 2023.4. The stress-free temperature condition was set to 165℃, and the boundary conditions were set to symmetry for 1 / 4 of the model and to cool the entire model from 165℃ to -65℃. Under these conditions, the simulation experiment was conducted, and the maximum stress generated in resin layer 5 at -65℃ was determined, which was taken as stress S1.
[0044] (2) Fracture toughness value S2 A film-like cured material with a thickness of 18 μm was formed by exposure and thermal curing of various photosensitive resin compositions. Strip-shaped test pieces with a width of 10 mm and a length of 30 mm were cut from the film-like cured material. Two notches extending from both ends of the test piece were formed at the center of the test piece along its length by laser irradiation using a laser scanning from both ends under the following conditions. The distance between the front ends of the two notches (ligament L) was set to approximately 1 mm, 2 mm, 2.5 mm, or 3 mm. A laser marking machine (Laser Workstation LW1, Greenenergy) manufactured by GRAVOTECH was used as the device for laser irradiation. Five test pieces each with ligament L of approximately 1 mm, 2 mm, 2.5 mm, or 3 mm were produced using the same method.
[0045] Marking speed: 100mm / s Wavelength: 532nm Marking frequency: 20kHz Marking power: 30% Tensile tests were conducted on notched specimens under the following conditions, and load-displacement curves were obtained from the crack propagation at the notch tip to the specimen fracture. A universal testing machine (AG-1kNXplus) with a specially shaped thermostatic bath manufactured by SHIMADZUCORPORATION was used as the apparatus for the tensile tests.
[0046] Stretching speed: 0.5mm / minute Spacing between fixtures: 20mm The fracture energy U accompanying the breakage of the specimen was calculated based on the area of the load-displacement curve obtained from the tensile test. The fracture energy U, ligament L, and specimen thickness t were then used to determine the fracture energy per unit area, U / Lt. The average value of all U / Lt values obtained from specimens with different ligament L values was taken as the fracture toughness value S2.
[0047] (3) Breaking stress S3 A 18 μm thick film was formed by exposure and thermal curing of various photosensitive resin compositions. Strip test pieces with a width of 10 mm and a length of 30 mm were cut from the film. Tensile tests were conducted on the obtained test pieces at 25 °C using a Shimadzu Corporation AGS-5kNG universal testing machine. The test speed was set to 5 mm / min, and the distance between the clamps was set to 20 mm. Flat clamping teeth (made of PET) were used as clamps. In the tensile test results, the breaking stress S3 was calculated by dividing the maximum load at which the test piece broke by the cross-sectional area of the test piece.
[0048] 2. Evaluation of electronic components On a printed circuit board with copper wiring, a resin layer with openings exposing the copper wiring is formed using various photosensitive resin compositions. The copper wiring exposed in the openings is then connected to a chip via solder balls. Next, an underfill portion is formed to fill the space between the chip and the resin layer, resulting in an electronic component for evaluation with a resin layer.
[0049] The obtained electronic components were subjected to reliability testing, which consisted of the following cycle: 15 minutes of rest at -65°C, followed by 15 minutes of rest at 150°C, repeated 2000 times. The chips and underfill of the electronic components after the reliability test were removed by grinding, exposing the resin layer. The surface of the resin layer was observed using a digital microscope (VHX-5000, manufactured by KEYENCE CORPORATION). Cracks in the solder mask (resin layer) were checked at 4000 locations near the copper wiring. The ratio of observed crack locations to the total number of observed locations was calculated as the crack initiation rate.
[0050] 3. Results The evaluation results are shown in Table 1. Table 1 also shows the values of the ratio S2 / S1 and the ratio S3 / S1.
[0051] [Table 1]
[0052] Table 2 shows the correlation coefficients between crack initiation rate and fracture toughness value S2, ratio S2 / S1, or ratio S3 / S1. The regression line representing the relationship between crack initiation rate and fracture toughness value S2 is expressed as crack initiation rate C[%] = -11.252 × S2 + 16.295. The regression line representing the relationship between crack initiation rate and ratio S2 / S1 is expressed as crack initiation rate C[%] = -919.73 × (S2 / S1) + 13.882. Based on these equations, crack initiation rate can be predicted based on the fracture toughness value S2 or ratio S2 / S1.
[0053] [Table 2]
[0054] Symbol Explanation 1-Circuit board, 3-Chip, 5-Resin layer, 5N-Notch, 7-Bottom filler, 7a-Rounded corner, 10-Substrate, 15-Metal wiring (copper pad), 20-Solder ball, 41, 42-Clamps, 50-Test piece, 100-Electronic component (analysis model), L-Ligament, Y-Distance between clamps, B-Width of test piece.
Claims
1. A method of predicting a generation rate of a crack in a resin layer provided adjacent to a metal wiring and formed of a resin material in an electronic part, comprising the steps of: predicting the generation rate of the crack from a fracture toughness value of the resin material.
2. A method of predicting a generation rate of a crack in a resin layer provided adjacent to a metal wiring and formed of a resin material in an electronic part, comprising the steps of: predicting the generation rate of the crack from a ratio S2 / S1 of a fracture toughness value S2 of the resin material to a stress S1 applied to the resin layer in the electronic part.
3. The method according to claim 2, further comprising the step of: calculating the stress S1 by a simulation test from a thermal expansion rate of the resin material.
4. A method of selecting the resin material in an electronic part provided with a metal wiring and a resin layer provided adjacent to the metal wiring and formed of a resin material, comprising the steps of: selecting the resin material from a fracture toughness value of the resin material.
5. A method of selecting the resin material in an electronic part provided with a metal wiring and a resin layer provided adjacent to the metal wiring and formed of a resin material, comprising the steps of: selecting the resin material from a ratio S2 / S1 of a fracture toughness value S2 of the resin material to a stress S1 applied to the resin layer in the electronic part.
6. The method according to claim 5, further comprising the step of: calculating the stress S1 by a simulation test from a thermal expansion rate of the resin material.
7. A method of manufacturing an electronic part provided with a metal wiring and a resin layer provided adjacent to the metal wiring and formed of a resin material, comprising the steps of: selecting the resin material by the method according to any one of claims 4 to 6; and providing the resin layer formed of the selected resin material adjacent to the metal wiring.
Citation Information
Patent Citations
Method for predicting incidence of cracks, method for selecting resin composition capable of suppressing occurrence of cracks, and method for manufacturing electronic component
JP2020188243A
Semiconductor module, electronic apparatus and printed circuit board
JP2021005586A
