Method for testing reliability of substrate for printed wiring board

Abstract

The present invention relates to a method for testing the reliability of a substrate for a printed wiring board. The method for testing the reliability of a substrate for a printed wiring board uses a nanoindenter to measure the load at which a crack occurs in the substrate for a printed wiring board.

Description

Technical Field

[0001] The present embodiment relates to a reliability test method for a printed wiring board substrate. Background Art

[0002] In recent years, the speed and capacity of signals used in electronic devices such as computers have been increasing, and the integration and functionality of semiconductor packages used in these electronic devices have also been increasing.

[0003] In addition, semiconductor packages are required to adapt to various temperature conditions depending on the equipment used, the installation environment, etc., and it is becoming increasingly important to ensure long-term reliability against temperature changes.

[0004] As a printed circuit board substrate, a substrate containing a resin component as an insulating material is used. However, due to the difference in thermal expansion coefficient between the resin component and the inorganic components such as the conductor layer and the semiconductor chip, stress is generated when the temperature changes, and cracks may occur in the printed circuit board substrate. In particular, in recent years, as the number of buildup layers has tended to increase, the stress caused by the difference in thermal expansion coefficient has gradually increased, and the problem of cracks caused by temperature changes has become prominent.

[0005] As a reliability test method for printed wiring board substrates, a temperature cycle test is generally performed (for example, refer to Patent Document 1).

[0006] Prior art literature

[0007] Patent Literature

[0008] Patent Document 1: Japanese Patent Application Publication No. 2006-093618 Summary of the invention

[0009] Problems to be solved by the invention

[0010] The temperature cycle test is an effective test method for confirming the long-term reliability of the test object against temperature changes by repeatedly heating and cooling. However, on the other hand, the temperature cycle test takes several weeks to months before the test results are obtained, so a simpler reliability evaluation method is desired.

[0011] In view of such a current situation, the present embodiment aims to provide a reliability test method for a printed wiring board substrate that can be easily performed.

[0012] Means for solving problems

[0013] The present inventors have conducted intensive studies to solve the above-mentioned problems, and as a result, have found that the above-mentioned problems can be solved by the present embodiment described below.

[0014] That is, this embodiment relates to the following [1] to [6].

[0015] [1] A reliability test method for a printed wiring board substrate, the reliability test method for a printed wiring board substrate comprising measuring a load at which cracks are generated in the printed wiring board substrate using a nanoindenter.

[0016] [2] The reliability test method for a printed wiring board substrate according to [1] above, wherein the printed wiring board substrate comprises a cured resin and glass cloth.

[0017] [3] The reliability test method for a printed wiring board substrate according to [1] or [2] above, wherein the printed wiring board substrate has two or more layers containing a cured resin and a glass cloth.

[0018] [4] The reliability test method for a printed wiring board substrate according to [2] above, wherein the indenter of the nanoindenter is pressed into the cured resin.

[0019] [5] A reliability test method for a printed wiring board substrate according to any one of [1] to [4] above, wherein the indenter of the nanoindenter is pressed into a cross section of the printed wiring board substrate.

[0020] [6] The reliability test method for a printed wiring board substrate according to any one of [1] to [5] above, wherein the indenter of the nanoindenter is a Glass indenter.

[0021] Effects of the Invention

[0022] According to the present embodiment, a reliability test method for a printed wiring board substrate that can be easily performed can be provided. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 is a graph used to illustrate the load-displacement curve.

[0024] Figure 2 is a graph illustrating the load-contact stiffness curve.

[0025] Figure 3 This is a schematic diagram for explaining the observation position using an optical microscope in the liquid tank temperature cycle test.

[0026] Figure 4 This is a schematic diagram for explaining the position of collecting the test piece and the test area in the destructive test using the nanoindenter.

[0027] Figure 5 This is a diagram for explaining the analysis method of the load-contact stiffness curve.

[0028] Figure 6 This is a graph showing the correlation between the crack initiation load W1 and the crack initiation rate after the temperature cycle test (1000 cycles) prepared from the results of the Examples. DETAILED DESCRIPTION

[0029] The action mechanism described in this specification is a guess, and does not limit the mechanism by which the effects of the present embodiment are achieved.

[0030] Any combination of the matters described in this specification is also included in this embodiment.

[0031] [Reliability test method for printed wiring board substrates]

[0032] The reliability test method of the printed wiring board substrate of the present embodiment is a reliability test method of the printed wiring board substrate for measuring the load at which cracks are generated in the printed wiring board substrate using a nano indenter.

[0033] According to the reliability test method of the present embodiment, since the test does not take a long time like the temperature cycle test, the reliability of the printed wiring board substrate can be easily evaluated.

[0034] <Nanoindenter>

[0035] The so-called nanoindenter used in the reliability test method of the present embodiment is a micro hardness tester that can press a micro indenter into the surface of a sample to a depth of several nm to several tens of μm and measure the relationship between the displacement and the load at that time.

[0036] <Substrates for printed wiring boards>

[0037] The printed wiring board substrate to be tested in the reliability test method of the present embodiment is not particularly limited, and a known printed wiring board substrate can be used as the test object.

[0038] The insulating layer of the printed wiring board substrate may be a layer containing resin but not glass cloth, or a layer containing resin and glass cloth. As the layer containing resin and glass cloth, a layer containing a resin cured product and glass cloth is preferred. It should be noted that in this specification, the so-called "resin cured product" means a cured product containing at least resin and no glass cloth, preferably a cured product of a thermosetting resin composition containing a thermosetting resin.

[0039] The thermosetting resin composition may contain a curing agent, a curing accelerator, an inorganic filler, and the like as components other than the thermosetting resin as necessary.

[0040] The insulating layer included in the printed wiring board substrate may be a single layer or a multilayer layer.

[0041] A specific example of a test object of the reliability test method of the present embodiment is a printed wiring board substrate having two or more layers including a cured resin and a glass cloth.

[0042] The printed wiring board substrate may have components other than the insulating layer that a printed wiring board substrate may have in general. Components other than the insulating layer include, for example, a conductor layer formed on the surface of the substrate, a conductor layer formed between the insulating layers, and a semiconductor chip mounted on the surface of the substrate.

[0043] The content of the cured resin in the layer containing the cured resin and the glass cloth is not particularly limited, and may be, for example, 25 to 75% by mass, 30 to 70% by mass, or 35 to 65% by mass.

[0044] The thickness of each layer containing the cured resin and the glass cloth is not particularly limited, and may be, for example, 200 to 1900 μm, 500 to 1800 μm, or 800 to 1700 μm.

[0045] In the case of a substrate having two or more layers containing a cured resin and glass cloth, the number of layers containing a cured resin and glass cloth is not particularly limited, and may be, for example, 2 to 19 layers, 5 to 18 layers, or 8 to 17 layers.

[0046] <Test conditions>

[0047] Next, the test conditions of the destructive test using the nanoindenter will be described.

[0048] In addition, in the following description, the test which measures the load which generates a crack in the substrate for printed wiring boards using a nanoindenter is referred to as a "destructive test."

[0049] In the following description, the load at which a crack is generated in a printed wiring board substrate measured in a destructive test is referred to as a "crack generation load."

[0050] The position of the printed circuit board substrate for the destructive test using the nanoindenter is not particularly limited, but for example, it is preferably selected at a location where cracks are likely to occur in a temperature cycle test. The location where cracks are likely to occur in a temperature cycle test is a location where stress is likely to occur in the printed circuit board substrate, so it is easy to obtain a crack generation load within the range of the destructive test using the nanoindenter.

[0051] When the printed wiring board substrate is viewed from above, as an example of a place for a destructive test, from the viewpoint of easily obtaining a crack-generating load, the end of the printed wiring board substrate is preferably used. In addition, from the viewpoint of easily obtaining a crack-generating load, the indenter of the nanoindenter is preferably pressed into the cross section of the printed wiring board substrate. It should be noted that since the printed wiring board substrate is a substrate obtained by cutting and processing the outer shape when viewed from above into a desired shape, the end face of the printed wiring board substrate is equivalent to the cross section.

[0052] When the printed wiring board substrate is viewed in cross section, an example of a location where the destructive test is performed is a substantially central portion in the thickness direction of the printed wiring board substrate.

[0053] That is, the location for performing the destructive test is preferably a cross section of an end portion of the printed wiring board substrate in a plan view, and is preferably a substantially central portion of the printed wiring board substrate in a cross-sectional view.

[0054] In addition, when performing a destructive test, it is preferable to process the printed wiring board substrate by cutting or the like to prepare a test piece including a region to be subjected to a destructive test in order to mount it on a nano indenter.

[0055] Regarding the surface of the test piece into which the indenter of the nanoindenter is pressed, from the viewpoint of suppressing the occurrence of deviations in the test results caused by the surface irregularities, it is preferably smoothed by grinding before or after the test piece is collected as a test piece. The grinding method is not particularly limited, and mechanical grinding or chemical mechanical grinding is preferred.

[0056] The test piece collected from the printed circuit board substrate can be pre-treated before the destructive test as needed. As an example of pre-treatment, heating treatment and cooling treatment can be cited. By implementing heating treatment and cooling treatment as pre-treatment, internal stress is accumulated in the test piece, and it is easy to obtain crack generation load in the scope of the destructive test using a nanoindenter.

[0057] (Destruction test conditions)

[0058] The type of the indenter of the nanoindenter is not particularly limited, but from the perspective of versatility, a Bosch indenter is preferred. The Bosch indenter is an indenter in which the shape of the surface of one side pressed into the test piece is a triangular pyramid (edge ​​spacing 115°). The type of the indenter and the edge spacing are not particularly limited, and a Bosch indenter (edge ​​spacing 100°), a Rockwell indenter, a Vickers indenter, a Knoop indenter or a Shore indenter can be used.

[0059] Since the indenter of the nanoindenter is very small, when a substrate including a plurality of materials such as a layer including a cured resin and glass cloth is used as a test object, the obtained load-displacement curve may change depending on the position of the indenter.

[0060] When the measurement object is a substrate having a layer containing a resin cured product and a glass cloth, if the position of the pressing head into the resin cured product is located at the glass cloth or a position close to the glass cloth, cracks are likely to occur in the glass cloth. If the resin cured product is located at a position moderately far from the glass cloth, cracks are likely to occur at the interface between the glass cloth and the resin cured product and cracks in the resin cured product. If the resin cured product is located further away from the glass cloth, cracks are likely to occur only in the resin cured product.

[0061] In a temperature cycle test of a substrate having a layer containing a cured resin and a glass cloth, cracks in the interface between the cured resin and the glass cloth and cracks in the cured resin are likely to become a problem, so it is preferred to adjust the position of the pressing head to a position where cracks in the interface between the cured resin and the glass cloth and cracks in the cured resin are likely to occur.

[0062] From the above viewpoints, it is preferred to press the indenter into the cured resin, preferably into the cured resin at a position moderately away from the glass cloth. On the surface of the indenter, the shortest distance between the glass cloth and the center of the indenter on the cured resin can be, for example, 1 to 10 μm, 2 to 8 μm, or 3 to 7 μm.

[0063] It should be noted that the position of the crimping tip can be adjusted using the optical microscope and micrometer attached to the nanoindenter.

[0064] The pressing speed of the indenter is not particularly limited, and may be, for example, 15 to 55 mN / s, 25 to 50 mN / s, or 35 to 45 mN / s.

[0065] The temperature for conducting the destructive test is not particularly limited, and may be 5 to 50°C, 10 to 40°C, or 20 to 30°C from the viewpoint of operability.

[0066] The atmosphere for conducting the destruction test is not particularly limited, but is preferably air from the viewpoint of operability.

[0067] (Analysis conditions)

[0068] The load-displacement curve is obtained using the above-mentioned destructive test.

[0069] An example of a load-displacement curve is shown in Figure 1 middle.

[0070] like Figure 1 As shown in the load-displacement curve shown in FIG, as the depth of the indenter into the test piece, that is, the displacement, increases, the load tends to increase overall. However, due to the cracks in the test piece, cracks such as Figure 1This is an area where the load increase becomes gentle, such as area A.

[0071] Will be used as Figure 1 The curve obtained by calculating the contact rigidity calculated from the slope of the load-displacement curve shown in FIG. 1 is shown in FIG. 1 and the curve obtained by calculating the load-displacement curve is shown in FIG. 1 (hereinafter also referred to as the “load-contact rigidity curve”). Figure 2 In. Figure 2 As shown, Figure 1 Region A in φ represents a downward peak where the contact rigidity decreases. Hereinafter, this peak is referred to as a "crack initiation peak."

[0072] As described in the examples described later, the load at which the above crack generation peak appears is correlated with the crack generation rate in the temperature cycle test. Therefore, the reliability of the substrate for the printed wiring board can be evaluated by measuring the load at which the crack generation peak appears (hereinafter also referred to as "crack generation load"). For example, the crack generation rate in the temperature cycle test can be predicted based on the crack generation load by pre-understanding the correlation between the above crack generation load and the crack generation rate in the temperature cycle test.

[0073] In the case of multiple crack generation peaks, a threshold value can be set for the size of the crack generation peak, and the crack generation peak with a predetermined size or more can be used as an indicator for reliability evaluation. The threshold value of the size of the crack generation peak can be set, for example, as the contact rigidity reduction rate (%) obtained by the following formula (1).

[0074] Contact rigidity reduction rate (%) = (S1-S2) × 100 / S1 (1)

[0075] (Wherein, S1 refers to the contact rigidity when the contact rigidity starts to decrease in the crack initiation peak, and S2 refers to the contact rigidity at the peak of the crack initiation peak.)

[0076] It should be noted that the above-mentioned "contact rigidity when contact rigidity starts to decrease" can be determined by the method described in the Examples, for example. In addition, the so-called "peak top" means the top of the crack generation peak (the point in the peak showing the minimum contact rigidity).

[0077] Once the crack initiation peak serving as an index for reliability evaluation is determined, for example, the load that provides the top of the crack initiation peak can be used for reliability evaluation.

[0078] When a plurality of crack initiation peaks satisfying the above threshold value appear, it is preferred to use the crack initiation peak that appears at the lowest load among them for reliability evaluation.

[0079] The crack generation peak suitable as an indicator of reliability may also change depending on the position of the indenter. Therefore, it is preferred to determine the position of the indenter, and under the condition of the indenter being pressed into the position, perform a temperature cycle test and a destructive test using a nanoindenter on a variety of test objects in advance to grasp the crack generation peak that has a high correlation with the temperature cycle test among the crack generation peaks that appear.

[0080] Example

[0081] The present embodiment will be described in detail below with reference to examples, but the present embodiment is not limited to the following examples.

[0082] [Manufacturing of printed wiring board substrates]

[0083] Production Examples 1 to 6

[0084] A laminated board formed by laminating 12 sheets of a predetermined insulating layer forming material (prepreg containing a thermosetting resin composition and glass cloth) was used as a core layer, and a laminate material (manufactured by Ajinomoto Co., Ltd., trade name "GX92", thickness 30 μm) (hereinafter also referred to as "Bu") and copper foil (thickness 18 μm or 12 μm) were used to produce printed wiring board substrates 1 to 6 (substrates with a thickness of 1.7 to 2.0 mm cut into a rectangular shape of 60 mm × 60 mm when viewed from above) having the following multilayer structure. It should be noted that the core layers of the obtained printed wiring board substrates 1 to 6 are different.

[0085] <Constitution of the substrate>

[0086] Copper foil (18μm) / Bu (30μm) / Bu (30μm) / Copper foil (18μm) / Bu (30μm) / Bu (30μm) / Copper foil (18μm) / Bu (30μm) / Bu (30μm) / Copper foil (18μm) / Bu (30μm) / Bu (30μm) / Copper foil (18μm) / Bu (30μm) / Bu (30μm) / Copper foil (12μm) / Core layer (1150~1350μm) / Copper foil (12μm) / Bu (30μm) / Bu (30μm) / Copper foil (18μm) / Bu (30μm) / Bu (30μm) / Copper foil (18μm) / Bu (30μm) / Bu (30μm) / Copper foil (18μm) / Bu (30μm) / Bu (30μm) / Copper foil (18μm) / Bu (30μm) / Bu (30μm)

[0087] It should be noted that the thickness in the above brackets means the thickness of each layer before lamination.

[0088] Then, a silicon semiconductor chip (with a rectangular shape of 20 mm×20 mm and a thickness of 0.775 mm when viewed from above) was arranged in the center of one surface of the printed wiring board substrate obtained above so that the circuit surface was on the lower side, and the gap between the semiconductor chip and the substrate was sealed using a liquid sealing material (Resonac Co., Ltd., product name "CEL-C-3730 series"), thereby obtaining substrates 1 to 6 with semiconductor chips for liquid tank temperature cycle tests.

[0089] [Liquid bath temperature cycle test]

[0090] In the liquid tank temperature cycle test, 5 substrates of each type with semiconductor chips mounted thereon were prepared, and the 5 substrates were subjected to a temperature cycle test in which they were repeatedly heated and cooled under the conditions of -65°C for 5 minutes, 150°C for 5 minutes, and a movement time between the low-temperature tank and the high-temperature tank of 18 seconds. After 1000 cycles, the substrates were taken out and the specified positions of the substrates were observed using an optical microscope to confirm whether cracks had occurred.

[0091] A perspective view for explaining the observation position using an optical microscope is shown in Figure 3 The observation position is Figure 3 In the illustrated substrate 1, eight locations are observed from two directions at four corners of the substrate indicated by regions 10 (hatched areas). The observation range is a region of 1500 μm×1500 μm per location.

[0092] The above 8 locations were observed for each of the 5 substrates, and the ratio of the locations where cracks were observed among the total 40 locations [number of locations where cracks were observed×100 / 40 locations] (unit: %) was defined as the crack generation rate in the bath temperature cycle test.

[0093] [Destructive test using nanoindenter]

[0094] Examples 1 to 6

[0095] 1. Preparation of test pieces

[0096] A perspective view for explaining the position of collecting the test piece and the position of performing the destructive test on the test piece is shown in FIG. Figure 4 middle.

[0097] The printed wiring board substrates 1 to 6 obtained in the manufacturing examples 1 to 6 were cut and collected. Figure 4 The test piece indicated by the portion 20 has a size of 1.5 mm (cut width W)×60 mm (substrate length L)×1.7 to 2.0 mm (substrate thickness T).

[0098] The test area for the destructive test in the collected test piece is Figure 4The region 30 is a cross section of the end of the printed wiring board substrate 1 and is approximately the center of the printed wiring board substrate 1 when viewed in cross section. In addition, the distance between the region 30 and the corner of the printed wiring board substrate 1 is ( Figure 4 The distance B) is about 10mm.

[0099] It should be noted that the cross section of the printed wiring board substrate including the above-mentioned region 30 was mechanically polished in advance using 4000-grade polishing paper. The obtained test piece was heat-treated at 260° C. for 6 hours in an air atmosphere, cooled to room temperature, and the obtained test piece was used as a measurement object of the nanoindenter.

[0100] 2. Destructive test using nanoindenter

[0101] The test piece obtained above was mounted on the nanoindenter in such a way that the indenter of the nanoindenter was pressed into the test area of ​​the test piece. Then, the position of the indenter was adjusted using an optical microscope and a micrometer attached to the nanoindenter so that the shortest distance between the center of the indenter on the surface of the indenter and the glass cloth on the cured resin was 5 μm, and a destructive test was performed under the following conditions.

[0102] (condition)

[0103] ■Device: Manufactured by Shimadzu Corporation, trade name "DUH-211S"

[0104] ■Type of indenter: Bosch indenter (edge ​​spacing 115°)

[0105] ■Pressing speed: 39mN / s

[0106] ■Maximum load: 490mN

[0107] ■Measurement temperature: 25℃

[0108] ■Measurement atmosphere: Air

[0109] 3. Analysis of measurement results

[0110] Among the crack initiation peaks appearing in the load-contact rigidity curve obtained by the above-mentioned destructive test, a crack initiation peak having a contact rigidity reduction rate of 37% or more calculated by the above-mentioned formula (1) was identified.

[0111] It should be noted that S1 in the above formula (1) is determined by the following steps. A schematic diagram for explaining the method of determining S1 is shown in Figure 5 middle.

[0112] The contact rigidity of the peak top of the peak to be analyzed is denoted as S2, and the load that provides S2 is denoted as W. S2 , WS2 -25mN is set as W 1A , W S2 -15mN is set as W 1B Then, in W 1A ~W 1B In the range of , the approximate straight line of the load-contact stiffness curve is obtained by the least square method and is set as “approximate straight line 1”.

[0113] Then, calculate the contact stiffness S2 and W S2 The absolute value D of the difference in contact rigidity of the approximate straight line 1 at the position S2 is set as W to provide a contact rigidity of S2+0.7D. 2A , the load that provides the contact rigidity of S2+0.5D is set to W 2B , in W 2A ~W 2B In the range of , the approximate straight line of the load-contact stiffness curve is obtained by the least square method and is set as “approximate straight line 2”.

[0114] The contact rigidity at the intersection of the approximate straight line 1 and the approximate straight line 2 obtained above is defined as S1.

[0115] Among the crack initiation peaks having a contact rigidity reduction rate of 37% or more determined by the above method, the crack initiation peak appearing at the lowest load is determined, and the load providing the peak top of the crack initiation peak is defined as the crack initiation load W1.

[0116] Figure 6 2 shows a graph in which the crack initiation load W1 obtained above is plotted on the horizontal axis and the crack initiation rate after the bath temperature cycle test (1000 cycles) is plotted on the vertical axis, and a linear approximation straight line of the above plot by the least square method.

[0117] from Figure 6 It is clear that the crack initiation load W1 and the crack initiation rate after the bath temperature cycle test (1000 cycles) have a high correlation, and it is found that the reliability test method of the present embodiment is useful as a reliability test method for printed wiring board substrates.

[0118] Description of Reference Numerals

[0119] 1 Substrate for printed wiring board, 10 Observation area, 20 Collection part of test piece, 30 Test area, W width, L length, T thickness, B distance, S1, S2 contact rigidity, W 1A , W 1B , W 2A , W 2B , W S2 Load, A area.

Claims

1. A reliability test method for a printed circuit board substrate, The reliability test method of the printed wiring board substrate measures the load at which cracks occur in the printed wiring board substrate using a nanoindenter.

2. The reliability test method for a printed wiring board substrate according to claim 1, wherein: The printed wiring board substrate includes a cured resin and glass cloth.

3. The reliability test method for a printed wiring board substrate according to claim 1 or 2, wherein: The printed wiring board substrate has two or more layers containing a cured resin and a glass cloth.

4. The reliability test method for a printed wiring board substrate according to claim 2, wherein: The indenter of the nanoindenter is pressed into the cured resin.

5. The reliability test method for a printed wiring board substrate according to claim 1 or 2, wherein: The indenter of the nanoindenter is pressed into the cross section of the printed wiring board substrate.

6. The reliability test method for a printed wiring board substrate according to claim 1 or 2, wherein: The type of the indenter of the nanoindenter is a Bosch indenter.

Images