JP3518545B2 - Display method of resin material - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for displaying a resin material, a container, a document, an adhesive strength (also referred to as peeling strength. The same applies hereinafter) measuring method and a composite manufacturing method. Regarding the display method of resin materials that can be separated and that can obtain universal adhesive strength that does not depend on the size and shape of the test piece, storage container, publication, adhesive strength measurement method and composite manufacturing method, and also the adhesive reliability The present invention relates to a resin material that can be easily evaluated.

[0002]

2. Description of the Related Art Electrons having a structure in which an insert member is resin-molded, such as a resin-sealed semiconductor device and a resin insulation transformer,
In an electric component, a high residual stress occurs at the resin bonding interface due to the curing shrinkage of the resin and the difference in linear expansion coefficient between the resin and the insert member.

Further, during the operation of these parts and the reliability test, higher heat stress is generated due to internal heat generation and severe heating and cooling, and peeling may occur at the adhesive interface.

Such interface peeling not only causes corrosion of semiconductor elements and electric wiring materials and deterioration of electric insulation, but also
Stress concentration due to peeling causes various damages such as resin cracks and fine wiring breaks.

Therefore, in order to ensure the reliability of such resin-molded parts, it is necessary to evaluate the adhesive strength of the resin material.

Conventionally, as a method for measuring the adhesive strength of a resin material, for example, IEE transaction
On Components Hybrids and Manufacturing Technology, Vol. 14, No. 4 (1991), pages 809 to 817 (IEEE
Trans. Comp. , Hybrids, Man
uf. Technol. , Vol. 14, No.
4 (1991) pp. 809-817) and adhesive technology,
Volume 9, Issue 1 (1990) pp. 60-63,
As described on pages 64 to 75 of the same magazine, a method is known in which a load such as pulling or shearing is applied to an adhesive test piece and the load when peeling occurs is divided by the adhesive area or the adhesive length. There is.

Further, a load is applied to an adhesive test piece partially provided with a peeling point, and the stress distribution near the boundary between the peeling portion and the bonding portion at the time of peeling progress, that is, the stress distribution near the boundary between the peeling portion and the bonding portion is uniquely described by the fracture mechanics parameter The method of doing so is known from the proceedings of the 67th Ordinary General Meeting of the Japan Society of Mechanical Engineers, Volume A (1990), pp. 75 to 77.

Further, a method of measuring the temperature at which peeling occurs due to residual stress in the cooling process after molding the adhesive test piece and then obtaining the residual stress distribution at the peeling starting point at that time by analysis is edited by the Japan Society of Mechanical Engineers, Volume A. , 54, 499 (1988), pages 597 to 603.

[Non-Patent Document 1] I-E-E Transaction on Components Hybrids and Manufacturing Technology, Volume 14,
Issue 4 (1991) pp. 809-817 (IE
EE Trans. Comp. , Hybrids,
Manuf. Technol. , Vol. 14,
No. 4 (1991) pp. 809-817)

[Non-Patent Document 2] Adhesion Technology, Volume 9, No. 1 (199
Year 0) Pages 60 to 63, Magazine pages 64 to 75
page

[Non-Patent Document 3] Proceedings of the 67th Ordinary General Meeting of the Japan Society of Mechanical Engineers, Volume A (1990), pages 75 to 77.

[Non-Patent Document 4] Transactions of the Japan Society of Mechanical Engineers, Volume A, 54th
Volume, 499 (1988) pp. 597-603
page

[0009]

Of the above-mentioned conventional techniques,
With the method of applying a load such as tension or shear, residual stress already exists at the time of making the adhesive test piece, so the measurable adhesive strength is only the apparent adhesive strength in which the residual stress is superimposed on the true adhesive strength. There is a problem that there is no.

Particularly, in the case of a transfer molding resin for semiconductor encapsulation, a release agent is mixed in the resin in order to facilitate the release of the resin from the molding die, and the adhesive strength is relatively low. Therefore, the rate of decrease in adhesive strength due to residual stress is large, and depending on the shape and size of the test piece, peeling may occur at the interface due to the residual stress alone.

Therefore, even if the adhesive strength measured by such a method is compared with the interfacial stress obtained by analysis or experiment, the adhesive reliability of the resin molded part cannot be evaluated.

Furthermore, the stress at the adhesive interface is not generally uniform, and in many cases, the stress has an infinite peculiarity at the end.

Since the distribution of stress and residual stress generated by the applied load during the adhesion test depends on the size, shape and material of the test piece, the load is divided by the adhesion area assuming a uniform stress distribution. Or, in the conventional method of dividing the load by the adhesive length assuming that the load acts only on the straight line along the peeling tip, the obtained adhesive strength depends on the size of the test piece,
You cannot get universal measurements.

When there is no residual stress, universal adhesive strength can be obtained by the method of describing the stress distribution in the vicinity of a singular point such as a peeling tip by a fracture mechanics parameter. However, when there is residual stress, the last example of the prior art (Journal of Japan Society of Mechanical Engineers, Volume A, Volume 54, Volume 4)
As in No. 99 (1988), pages 597 to 603), it is necessary to obtain a residual stress distribution by analysis. When the residual stress is obtained by analysis, depending on the material of the resin, the temperature dependence of the physical properties and the viscoelastic behavior at high temperatures are significant, so the analysis may be extremely complicated or difficult to perform with high accuracy.

There is also a problem that the adhesive strength cannot be freely measured at any necessary temperature when peeling is caused only by residual stress as in the last example of the prior art described above.

As described above, since a universal adhesive strength as a physical property value has not been practically obtained in the past, it is difficult to quantitatively predict the adhesive reliability at the interface of a resin molded product. It was necessary to actually make a molded product and inspect and measure the adhesive state and adhesive strength at the interface.

A first object of the present invention is to separate a true adhesive strength from a residual stress and to measure a universal adhesive strength which does not depend on the size and shape of a test piece with high accuracy and easily. To provide a method for displaying a resin material, a storage container, a written document, and a method for producing a composite using the result.

A second object of the present invention is to provide a resin material capable of easily predicting and evaluating the adhesion reliability of the interface of the molded product from the given physical property values.

[0019]

The first object of the present invention is to have a partial peeling point between a resin material and a mating material (hereinafter referred to as an adherend) for which the adhesive strength between the resin material and the resin material is required. , Create a bonding test piece with a shape such that the residual stress at the bonding interface near the peeling tip is mainly the shear stress component, and apply a load such that normal and reverse shear stress acts on the bonding interface near the peeling tip. Is individually loaded to obtain the apparent peeling strength in each case.

The second object can be achieved by attaching to the resin material the measurement result of the adhesive strength obtained by the above means.

In the method for displaying a resin material, the storage container and the publication of the present invention, the adhesive strength based on the shear stress of the resin to the adherend should be described together with the name of the adherend, or substantially the influence of residual stress. It is characterized in that the adhesive strength of the resin to the adherend obtained by excluding the above is described.

In this case, the resin material is not molded (before polymerization, etc.).
In the state of (1) (liquid, powder, etc.), it means a resin composition particularly in the case of a thermosetting resin. Further, the resin means a resin cured product. The notation is preferably indicated by a stress intensity factor or strain energy release rate.

The stress intensity factor is generally expressed in units of MPa√m or kgf / mm ^{3/2 (3} square of 2 minutes), 0.5 square of [stress] × [length], [force] × [Length] -1.
It is represented by the dimension of the 5th power, or [mass] × [length] −0.5 power × [time] −2 power.

The strain energy release rate is expressed in the unit of J / m ² or kgf / mm, and is [energy] × [length].
It is expressed in the dimensions of the second power of −, [force] × [length] to the first power, and [mass] × [time] to the second power.

The method for measuring the adhesive strength of the resin material of the present invention is as follows:
In a structure composed of two or more materials containing at least one resin, a partial peeling point is previously formed between two materials adhered to each other (that is, at an adhesive interface between the at least one resin and another material). Two kinds of loads are applied to the bonding interface so that opposite shear stresses act on each other, and the peeling progress strength for each load is determined, or two materials that determine the mutual bonding strength are used. By applying a bending load in the direction perpendicular to the bonding interface by laminating in layers and applying a reverse shearing stress by reversing the direction, or between two materials bonded together,
It is characterized in that the residual stress acting on the adhesive interface and the stress in the same direction and in the opposite direction are individually applied to obtain the adhesive strength.

According to the method for producing a composite of the present invention, when forming a composite of a resin and a metal, the combination of the resin and the metal is based on the adhesive strength of the resin obtained by using the metal as an adherend. It is characterized by selecting.

According to the method of manufacturing a semiconductor device of the present invention, when the metal lead frame and the semiconductor element are molded with the sealing resin, the sealing resin obtained by using the material of the metal lead frame as the adherend is adhered. One of the material of the sealing resin, the material of the metallic lead frame, the surface treatment condition of the metallic lead frame, or a combination thereof is selected based on the strength.

In these cases, the resin is epoxy based,
The metal is preferably selected from copper, copper alloys, iron, aluminum or alloys thereof such as iron-42 nickel. However, the resin is not limited to this, and may be thermoplastic or thermosetting. Further, the resin material may be liquid or powder, and it does not matter whether the resin material is cured by heat or not. The adherend is not limited to metal, and the present invention can be applied to either ceramics or resin.
As the cured product, any of a film, a plate-shaped product and a bulk product can be adopted depending on the application.

Examples of the resin include thermosetting resins such as epoxy resins, silicone resins, and phenol resins, and thermoplastic resins such as polyethylene resins and polyamide resins, which may contain additives. It is also possible to use an adhesive as a resin, for example, a thermosetting resin such as an epoxy resin base, a thermoplastic resin such as a vinyl acetate resin base, an elastomer such as a chloroprene base, a mixed type of a phenol resin-epoxy resin, etc. Resin etc. are mentioned.

Further, the result of the strength display according to the present invention is particularly effective when applied to a composite body in which adhesion between resin and metal is required, and electronic parts such as resin-sealed semiconductor devices,
Suitable for power equipment such as resin insulation transformers and home electric appliances such as VTR chassis.

In the present specification, the peeling strength means the strength against further peeling starting from the portion peeled in advance.

The display does not prevent the simultaneous writing with other conditions.

By applying two kinds of forward and reverse shear stresses, it is possible to obtain two peeling progress strengths in which the residual stress increases or decreases the stress due to the applied load. The apparent stress distribution near the exfoliation tip, which is generated only by the applied load, can be calculated accurately from the dimensions, shape, and physical properties of the test piece, so by taking the arithmetic mean of these two apparent strengths, A universal true bond strength can be obtained.

That is, since there is always residual stress in the joint, "true strength + residual stress" is measured in the measurement in which a shear stress in the direction opposite to the residual stress is applied to the adhesive interface. , In the measurement in which the shear stress acts in the same direction as the residual stress, "true strength-residual stress" is measured, so the average of both measurement results, that is, "
(True strength + residual stress) + (true strength-residual stress) "/
This is based on the principle found by the present inventor that the true strength can be found by finding 2.

In addition, if such a measurement result of the adhesive strength is given to the resin material, the stress generated in the molding state based on the physical properties such as the longitudinal elastic modulus and the linear expansion coefficient which have been conventionally used. Can be obtained by analysis, and by comparing this with the adhesive strength, the adhesive reliability of the molded product can be quantitatively predicted and evaluated.

[0035]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a front view showing a shape of a test piece and a load application method in an adhesive strength measuring method for a resin material according to an embodiment of the present invention.

The strip-shaped or rectangular parallelepiped test piece 1 has a two-layer structure in which a resin 2 and an adherend 3 are adhered to each other, and a peeling portion 5 is previously provided at one end of an adhesion interface 4. Specimens of this shape subjected to bending loads are generally ENF
(End-Notched Flexure) Test piece.

The resin material used in this example uses cresol novolac epoxy resin as the main component, phenol novolac resin as the curing agent, and fused silica as the filler. In addition to this, a plasticizer, a curing accelerator, and a coupling agent are used. Agent, release agent,
It is an epoxy resin composition to which a flame retardant, a coloring agent and the like are added in a slight amount.

In producing the test piece 1, the adherend 3 is put in the mold in advance, the resin 2 is molded and cured at a high temperature, and then cooled to room temperature or heated to the test temperature. Cooled.

Therefore, the residual stress τ _r having a shear stress component as a main component acts on the adhesive interface 4 due to the curing shrinkage of the resin 2 and the difference in linear expansion coefficient from the adherend 3. In FIG. 1, the action direction of the residual stress τ _{r in} the vicinity of the peeling tip 7 is shown as an example in which the shrinkage of the resin 2 is larger than the shrinkage of the adherend 3.

The adhesive strength test is a three-point bending test for both cases in which the test piece 1 is turned upside down as shown in FIGS. 1 (a) and 1 (b), respectively. The load was measured. That is, in FIG. 1A, the test piece 1 was supported with the two fulcrums 6 with the resin 2 side facing upward, and the load P ₁ was applied to the center of the span. Further, in FIG. 1B, a load P ₂ was similarly applied with the adherend 3 side facing up.

[0042] In the vicinity of ablation tips 7 of the bonding interface 4 at this time, shear stress tau ₁ by the load P _1, the shear stress tau ₂ due to the load P ₂ are respectively in FIG. 1 (a), in the direction indicated in (b) To work.

Regarding the examples shown in FIGS. 1A and 1B, the relationship between the stresses at the start of delamination is schematically shown in FIGS. 2A and 2B, respectively.

In FIGS. 2A and 2B, the absolute values of the residual stress τ _r are the same in both cases (a) and (b), and the sign is reversed by simply reversing the coordinate system. In the case of (a), as the load P ₁ is applied, the shear stress τ due to the load P _1
1 first reduces the residual stress τ _r , then reverses the sign,
When the sum of both stresses τ _r and τ ₁ reaches the critical shear stress τ _c , which is the true adhesive strength, the progress of delamination starts.

On the other hand, since in the case of (b) acts in a direction of increasing shear stress tau ₂ due to the load P ₂ further residual stress tau _r, a small load P ₂ than the load P ₁ in the case of (a) The stress sum reaches the critical shear stress τ _c , and the peeling progresses.

In FIGS. 1 and 2, the test piece 1 is turned upside down with the load directions of the loads P ₁ and P _{2 being} the same.
Considering that the coordinate system is defined with the test piece 1 as a reference and the direction of load application is reversed, the residual stress τ _r between the directions (a) and (b) in FIGS. ₁ and τ ₂ are in opposite directions.

FIG. 3 shows an adhesion test piece using a semiconductor encapsulating epoxy resin and a semiconductor lead frame Fe-42Ni alloy plate, and a load P ₁ was applied from the resin 2 side as shown in FIG. 1 (a). In this case, measure the peeling start load,
It is the result of analyzing the stress distribution near the peeling tip 7 along the adhesive interface under this load by the finite element method.

Since the Fe-42Ni alloy has an extremely small coefficient of linear expansion as compared with general resin materials, it has a large residual stress during the preparation of test pieces, and conventionally it has been particularly difficult to quantitatively measure the adhesive strength with the resin material. It is a material.

The dimensions of the test piece are 55 mm in length and 6 m in width.
m, the thickness of the resin and the Fe-42Ni plate are 1.5 mm and 0.25 mm, respectively, and the fulcrum spacing in the three-point bending test is 45 mm, and the distance from the fulcrum on the side where the peeling portion is provided to the peeling tip is 10
mm, molded and cured at 175 ° C. and tested at room temperature.

In the stress analysis, it is assumed that the residual stress is generated by cooling from the mold temperature to room temperature, and the thermal stress and the load are taken into consideration.

In FIG. 3, when the coordinate system of the test piece 1 is the x-axis in the direction parallel to the adhesive interface 4 and the y-axis in the direction perpendicular to the adhesive interface 4, the vertical stress σ which is two stress components involved in the progress of peeling. The distributions of _y and shear stress τ _xy are shown.

As shown in the figure, the shear stress τ _xy is much larger than the normal stress σ _y in the vicinity of the peeling tip 7,
In the measuring method of this example, it can be seen that the progress of peeling is largely controlled by the shear stress component.

FIG. 4 shows the same test piece as in FIG.
The peeling initiation load for the two types of load shown in Fig. 2 is measured, and the distributions of the apparent shear stresses τ ₁ and τ ₂ generated near the peeling tip due to only these loads and the shear stress τ _r due to the thermal stress This is the result of analyzing the distribution by the finite element method.

Further, in FIG. 4, (τ ₁ + τ ₂ ) / 2 and (τ
The distribution of _{1 −} τ ₂ ) / 2 and the distribution of the critical stress τ _c at the initiation of delamination corresponding to τ _xy in FIG. 3 are also shown. As can be seen from FIG. 4, the arithmetic mean (τ ₁ + τ ₂ ) / 2 of the apparent shear stresses τ ₁ and τ ₂ due to only the applied load is in good agreement with the critical stress τ _c at the start of delamination growth, and (τ ₁ −τ ₂ ) / 2 is consistent with the shear stress τ _r due to thermal stress only.

Therefore, as shown in FIG. 1, the apparent shear stresses τ ₁ and τ _{2 in} both cases where the test piece 1 was turned upside down.
It can be seen that the critical shear stress τ _c , which is the true adhesive strength, and the residual stress τ _r can be separated by determining

The limit shear stress τ obtained as described above
_Since the distribution of _c is not a single numerical value, it is inconvenient to use it as a true adhesive strength as it is. So Fig. 3
Fracture mechanics parameters such as a stress intensity factor and a strain energy release rate are used as the parameters expressing the stress distribution at the peeling tip as shown in FIG.

The stress distribution on the adhesive interface near the peeling tip is
The stress intensity factors K _I and K _II for deformation of the aperture type (mode I) and the in-plane shearing type (mode II) are expressed by the following equation.

[0058]

[Equation 1]

[0059]

[Equation 2]

[0060]

[Equation 3]

Here, σ _y , τ _xy and r are the normal stress, the shear stress and the distance from the exfoliation tip shown in FIG. 3, respectively, and π.
Is the pi, i is the imaginary unit, and d is the representative length. μ and υ
Are respectively the lateral elastic modulus and the Poisson's ratio of the material, and the subscripts p and a distinguish between the resin and the adherend. By combining the two parameters of K _I and K _II as expressed by Equation 1, it is possible to express the limit stress distribution corresponding to the true adhesive strength.

In the case of the interface of different materials, as expressed by the equation 1, K _I and K are different from those of the crack in the homogeneous material.
_{Since II} does not correspond to σ _y and τ _xy individually, they cannot be considered separately.

Therefore, even when the shear stress component is dominant as in this embodiment, the stress distribution cannot be expressed only by K _II , and it is necessary to use a combination of K _I and K _II .

As one of the methods of expressing the stress distribution near the exfoliation tip with a single parameter, the method of using the stress intensity factor K _i expressed by the following equation when no residual stress exists is a term of the prior art. It is known from the papers published in the proceedings of the 67th Ordinary General Meeting of the Japan Society of Mechanical Engineers shown in.

[0065]

[Equation 4]

This parameter can also be applied to the separation between the true adhesive strength and the residual stress described in the above embodiment, as compared with the case where the combination of K _I and K _II is used, as shown below. This has the advantage of facilitating parameter calculation and evaluation.

That is, in the above embodiment, since τ _xy >> σ _y as shown in FIG. 3, K _i can be expressed by the following equation, which has a one-to-one correspondence with the distribution of shear stress τ _xy. Will correspond.

[0068]

[Equation 5]

Therefore, when the test piece is turned upside down as shown in FIG. 1 and the apparent K _i (hereinafter, respectively represented by K _i1 and K _i2 ) due to only the applied loads P ₁ and P ₂ are obtained, the arithmetic mean of these is calculated. From (K _i1 + K _i2 ) / 2 to K corresponding to the true adhesive strength
_i (hereinafter referred to as K _ic ) is also (K _i1 −K _i2 ) / 2.
From this, K _i (hereinafter referred to as K _ir ) corresponding to the residual stress can be obtained, and the adhesive strength can be evaluated using only a single numerical value.

To obtain K _i for an arbitrary load condition, from the stress distribution on the adhesive interface near the peeling tip or the displacement distribution of the peeling surface analyzed by a numerical analysis method such as the finite element method or the boundary element method, Proceedings of the Japan Society of Mechanical Engineers, Volume A, Volume 55, No. 510 (1989), pages 340 to 34.
It may be calculated by the method described on page 7.

Further, K _{i with} respect to the load applied in the embodiment of FIG.
Like the strain energy release rate described later, can also be easily calculated from the beam bending theory without performing a numerical analysis.

FIG. 5 shows the apparent stress intensity factors K _i1 and K _i2 due to only the applied load and the stress intensity factor K _ir due to only the thermal stress, the load and the heat, with respect to the analysis result of the finite element method shown in FIG. It is the result of calculating the stress intensity factor K _ic corresponding to the critical stress at the start of delamination development in consideration of both stresses.

Although the numerical analysis error is large in the vicinity of the exfoliation tip, almost constant values are obtained for the respective stress intensity factors in the other regions.

As apparent from the figure, (K _i1 + K _i2 ) / 2 and (K _i1 −K) obtained from the apparent K _{i with} respect to the load load.
It can be seen that _i 2) / 2 is in good agreement with the true adhesive strength and the stress intensity factors corresponding to the residual stress, K _ic and K _ir , respectively.

In the above example, the residual stress and the true adhesive strength are obtained by performing thermal stress analysis. However, depending on the material of the resin, the temperature dependence of the physical properties is high, so detailed temperature-dependent data such as the coefficient of linear expansion and the coefficient of longitudinal elasticity, which correspond to the thermal history from the test piece mold to the adhesive strength test, are analyzed prior to analysis. Is necessary, and the viscoelastic behavior at high temperature is remarkable, so the analysis is extremely complicated and the residual stress analysis with high accuracy may be difficult.

According to the method of this embodiment in which the residual stress is separated and eliminated only from the apparent K _i against the applied load, the true adhesive strength can be obtained easily and accurately without analyzing the residual stress. it can.

Since the stress intensity factor K _i can uniquely describe the strength of stress in the vicinity of the peeling tip, it is possible to obtain a universal adhesive strength independent of the size and shape of the test piece.

K _{i with} respect to the load applied in the embodiment of FIG.
The bending theory of beams can be used to derive the following.
That is, the width of the test piece 1 b, the thickness of the respective t _p of the resin 2 and the adherend 3, the distance t _a, 3-point bending fulcrums 6 2L, from the side of the fulcrum 6 provided the peeled portion 5 The peeling length up to the peeling tip 7 is a, and the longitudinal elastic moduli of the resin 2 and the adherend 3 are E _p and E _{a, respectively.}
Then, the deflection δ of the load point when the load P is applied to the center between the fulcrums 6 is obtained from the bending theory of the beam as follows.

[0079]

[Equation 6]

Since the compliance C with respect to the applied load P is δ / P, the strain energy release rate G is as follows.

[0081]

[Equation 7]

The stress intensity factor K _i and the strain energy release rate G have the following relationship.

[0083]

[Equation 8]

Therefore, the stress intensity factor K _i can be easily obtained from the equations 7 and 8 by arithmetic calculation.

The longitudinal elastic modulus E used in the above calculation,
Since the three material constants of the lateral elastic coefficient μ and the Poisson's ratio υ are related by the following equation, only two of them need be obtained in advance in the calculation.

[0086]

[Equation 9]

The physical property values necessary for the analysis of the apparent stress or the stress intensity factor only by the load are the values of these physical property values at the test temperature regardless of the thermal history from the mold temperature. The strain energy release rate G can be used as a parameter representing the true adhesive strength instead of the stress intensity factor K _i .

In this case, the strain energy release rate G is the stress
Expansion factor K_iAnd proportional to the square of the stress, so negative
Load P₁, P₂Strain energy release rate for
G₁, G₂To the true adhesive strength and residual stress.
Energy release rate G_c, G _rWhen calculating
It is necessary to add and subtract to the square root.

[0089]

[Equation 10]

[0090]

[Equation 11]

As a parameter showing the true adhesive strength,
In addition to the stress intensity factors K _I , K _{I I} , stress intensity factor K _i , and strain energy release rate G described above, the stress intensity near the exfoliation tip is uniquely described regardless of the size and shape of the test piece. Any parameter that can be used, such as the path-independent integral J used in fracture mechanics, can be used.

Further, practically, after the size and shape of the test piece are specified, the value of the peeling progress start load due to the load load from the normal and reverse directions, and the forward and reverse directions with the same peeling length are measured. You may use the value of the average peeling start load of.

Even when these load values are used, they can be converted into arbitrary universal parameters as needed at any time.

The resin material for measuring the adhesive strength may be either thermosetting or thermoplastic. Further, the material of the adherend may be metal, various inorganic materials such as ceramics, silicon, glass, or other resin materials, or the same resin material molded separately. The types of residual stress are not only thermal stress but also shrinkage due to curing reaction, moisture,
It is possible to separate residual stress caused by any cause such as swelling due to permeation of a chemical solution or the like, material change due to physical or chemical environment, and the like.

In determining the dimensions of the test piece 1,
Care must be taken to meet the following conditions.

That is, (1) the adhesive interface 4 is not peeled due to residual stress in the stage of producing the test piece 1, (2) the resin 2 or the adherend 3 is not broken or plastically deformed before the peeling of the adhesive interface, (3) During the bending test, large deformation that does not fall within the applicable range of beam bending theory and linear numerical analysis does not occur.
(4) The effect of shear stress acting in the width direction of the test piece 1, that is, the direction perpendicular to the paper surface of FIG. 1, can be ignored.

Regarding the condition (1), it is desirable that the resin 2 is as thin as possible in comparison with the length and width of the test piece,
Regarding conditions (2) and (3), on the contrary, it is desirable that the resin 2 is not extremely thin as compared with the fulcrum spacing for three-point bending.
Regarding the condition (4), it is desirable that the width of the test piece 1 is sufficiently smaller than the fulcrum spacing.

The limit value of each dimension satisfying the above conditions varies depending on the combination of the materials of the resin 2 and the adherend 3, but the following is a rough guideline.

That is, the thickness of the test piece is in the range of about 1/5 to 1/40 of the fulcrum interval, and the width of the test piece is 5 of the fulcrum interval.
It is desirable that it is 1 / or less.

The peeling tip 7 is preferably separated from both the fulcrum 6 and the load point by a distance equal to or larger than the thickness of the test piece.

To provide the peeling portion 5 at one end of the test piece 1 prior to the adhesive strength test of FIG. 1, a mold release agent is applied to one end of the adherend 3 before the test piece 1 is molded, or fluorine is applied. A tape made of a material with poor adhesion such as resin may be attached.

Even when these peeling means are not used before molding, the interval between the fulcrums 6 is narrowed before the adhesive strength test, and a local bending load is applied near the end of the test piece 1, The peeled portion can be formed by pressing a razor blade against the end of the test piece 1.

Even when a peeling means is used before the molding, it is better to further promote natural peeling from the tip of the obtained peeled portion by bending load or the like to obtain a more accurate adhesive strength without the influence of the tape or the like. Can be measured.

The length of the peeled portion 5 necessary for calculating the stress intensity factor K _i and the like is measured by observing the side surface of the test piece with a microscope, ultrasonic flaw detection from the upper and lower surfaces, or the like. When the peeling means is used before the test piece is molded, the coating length of the release agent or the adhesive length of the tape may be measured. However, in the latter case,
It is premised that peeling does not proceed after molding.

The start of peeling progress during the three-point bending test can be detected by the change in compliance C due to the change in peel length, that is, the bending of the curve showing the relationship between the load P and the deflection δ. Further, an acoustic emission detector, a microphone, or the like may be used to detect the acoustic signal generated when the peeling progresses.

In the above description, the adhesive strength was obtained based on the starting point of peeling progress. However, depending on the material, the stress intensity factor K _i, etc. at the beginning of peeling progress, during progress, and at the time of progress stop may differ. . In such a case, a value at an appropriate time may be adopted according to the intended use of the adhesive strength.

FIGS. 6 and 7 are front views showing the shape of the test piece and the load application method in the method for measuring the adhesive strength of the resin material, which is another embodiment of the present invention.

The test pieces used in the adhesive strength measuring method of the present invention are not limited to the resin 2 and the adherend 3 for which the mutual adhesive strength is required.
It is not necessary to be composed of only the above two materials. For example, as shown in FIG. 6, the adhesive strength of the resin 2 sandwiched between the same or different adherends 3a and 3b at the adhesive interface 4 on the adherend 3a side is measured, or As shown in 3
It is also possible to measure the adhesive strength between the resin 2 and the second adherend 3a provided on the surface of b by plating, painting, adhesion, vapor deposition or the like.

Furthermore, not only the two-layer and three-layer structure as shown in FIGS. 1, 6, and 7, but also the multilayer structure of four layers or more, and the specific material is present only in a part of the entire length of the test piece. May be. In these cases, the peeled portion 5 needs to be provided on the adhesive interface 4 between the materials whose adhesive strength is to be measured.

In FIGS. 6 and 7, the load application method is shown for the case where the load P ₁ is applied from only one direction, respectively. Separating the adhesive strength and the residual stress is the same as in the embodiment of FIG.

FIG. 8 is a front view showing the shape of the test piece and the method of applying a load in the method for measuring the adhesive strength of a resin material which is still another embodiment of the present invention. In this embodiment, in the direction parallel to the bonding interface 4 of the test piece 1 in which the resin 2 and the adherend 3 are bonded, compression load as shown in FIG. 8A and tensile load as shown in FIG. 8B are applied. By loading P ₁ and P ₂ ,
Shear stresses τ ₁ and τ ₂ that are reversed in the normal and reverse directions are applied.

As described above, according to the adhesive strength measuring method of the present invention, the shear stress in the vicinity of the peeling tip 7 can be reversed in both the forward and reverse directions, and its magnitude is larger than that of the normal stress in the direction perpendicular to the adhesive interface 4. Any shape of test piece and any loading method can be used as long as it can be made sufficiently large.

When a load in the compressive and tensile directions is applied as shown in FIG. 8, in order to reduce the vertical stress component in the direction perpendicular to the adhesive interface 4, the load applied in opposition to both ends of the test piece 1 is required. It is necessary to take care that the bending moment does not act on the test piece 1 because they are on the same axis.

When the adhesive strength test is conducted by bending load, in addition to the three-point bending load shown in FIG. 1 and the like, four-point bending load and bending load is applied to a test piece supported in a cantilever shape. Can be used, and the stress intensity factor K _i and the strain energy release rate G at that time can be derived from the beam bending theory.

FIG. 9 is an example of a document describing the characteristics of the resin material, which describes the measurement results of the adhesive strength by the method of the present invention.

With respect to both the resin material and the adherend, if physical properties such as a longitudinal elastic modulus conventionally used, a bending elastic modulus, a Poisson's ratio, a linear expansion coefficient, etc., which are used instead of the elastic modulus, are measured or given, the From the value of, the stress generated at the adhesive interface inside the resin mold component or the like in the molded state can be predicted by analysis.

Therefore, if the true adhesive strength based on the parameters such as K _i obtained by the method of the present invention is given in the form of the document shown in FIG. By comparing, it is possible to quantitatively predict and evaluate the presence or absence of peeling at the interface and the degree of peeling, without actually forming a molded product of a resin material.

In the adhesive strength measuring method of the present invention, the strength against peeling progress when the peeled portion already exists is calculated. Therefore, when the presence or absence of peeling of a molded product is predicted based on the resulting adhesive strength. In the case of, it is sufficient to evaluate the presence or absence of delamination development from the presence of a minute delamination portion.

Since the conventionally used adhesive strength depends on the size and shape of the test piece, it can be used only for relative comparison between materials, but the adhesive strength measured by the method of the present invention is Since it can be applied to the quantitative evaluation of the adhesion reliability of, in the notation of the measurement result, a method of upside down three-point bending test or a method of reversing shear stress, etc.
It is desirable to display the measurement method.

The measurement result of the adhesive strength is not limited to the inspection report shown in FIG. 9, but the effect that enables the prediction and evaluation of the adhesive reliability of the resin material even if it is described in various specifications of the resin material or in the container There is.

In the analysis of the generated stress, the Poisson's ratio of the resin material is necessary as described above. However, the Poisson's ratio is more complicated to measure than the flexural modulus and the stress analysis result is not affected. Since it is small, the description may be omitted as in the example of FIG.

Since the adhesive strength of the resin material changes depending on the material and surface condition of the adherend, environmental conditions such as temperature and humidity, molding conditions, etc., when describing the adhesive strength measurement results,
It is desirable to write these measurement conditions together with the numerical values of the measurement results, or display them in the form of a graph for the measurement conditions such as temperature.

FIG. 10 shows the epoxy resin for semiconductor encapsulation and the Fe for semiconductor lead frame obtained by the method of the present invention.
It is a graph showing the relationship between the true bond strength K _ics and temperature -42Ni alloy plate.

In the conventional adhesive strength measuring method, not only the true adhesive strength but also the residual stress changes with temperature, so that the temperature dependence of the adhesive strength cannot be quantitatively obtained. If the true bond strength is given as a function of temperature as shown in FIG. 10, it is also possible to predict the critical temperature at which peeling occurs by comparison with the stress generated in the molded product obtained by analysis as a function of temperature.

Next, an example of predicting the peeling occurrence temperature of the adhesive interface between the lead frame and the sealing resin when heating the resin-sealed semiconductor device based on the adhesive strength by the method of the present invention will be shown. FIG. 11 is a cross-sectional view showing the structure of the resin-encapsulated semiconductor device which is the object of evaluation, and FIG. 12 is a prediction result of the peeling occurrence temperature.

In FIG. 11, the semiconductor element 8 is made of Fe-
The lead frame is made of 42Ni alloy, and is fixed to the tab 9 of the lead frame with an adhesive or the like. A plurality of leads 10 are formed around the tab 9 with the same lead frame material. The electrodes on the surface of the semiconductor element 8 and the leads 10 are electrically connected by a thin metal wire (not shown),
Each of these members is molded with an epoxy-based sealing resin 11 except for the lead-out portion of the lead 10.

The resin-encapsulated semiconductor device as shown in FIG. 11 is exposed to a high temperature of 200 ° C. or higher when it is mounted on a wiring board by soldering, and the thermal stress at this time causes the bonding interface of each part in the semiconductor device. Peeling may occur. Therefore, it is assumed that a minute peeled portion 12 as shown in FIG. 11 exists at the end of the adhesive interface with the sealing resin 11 on the lower surface of the tab 9,
The value of the stress intensity factor K _i generated at the peeling tip 13 when the resin-sealed semiconductor was heated to various temperatures was analyzed by the finite element method. Furthermore, using the method of the present invention, the true adhesive strength K _ic between the sealing resin 11 and the lead frame material at various temperatures was measured and compared with the above analysis results.

FIG. 12 shows the comparison result. In FIG. 12, the curve rising to the right shows the relationship between the stress intensity factor K _i generated by the analysis and the heating temperature, and the curve descending to the right shows the relationship between the true adhesive strength K _{ic found} by the experiment and the measurement temperature. The intersection of both curves gives the peeling generation temperature, that is, the peeling development generation temperature from the minute peeling portion 12. Further, on the curve showing the analysis result of the stress intensity factor K _i of FIG. 12, actual resin-sealed semiconductor devices were left in a constant temperature bath at various temperatures for 10 minutes, and then tabbed by an ultrasonic inspection apparatus. 9 The results of observing the occurrence of peeling on the lower surface are shown by various symbols.

White circles represent samples in which peeling was not observed on the lower surface of the tab 9 at that temperature, and black circles represent samples in which the entire lower surface of the tab 9 was peeled at that temperature. The circles in which black is mixed indicate the magnitude of partial peeling near the lower end of the tab 9 depending on the magnitude of the black area.

[0130] As seen in FIG. 12, the peeling of the tab 9 underside generated stress intensity factor K _i is at a lower temperature region than the true bond strength K _ics is not observed, the entire surface is at a higher temperature range than the true bond strength K _ics Peeling and partial peeling were observed in the temperature region near the intersection of both curves.

This result shows that the occurrence of peeling inside the resin-sealed semiconductor device can be predicted by comparing the adhesive strength obtained by the method of the present invention and the stress analysis result of the resin-sealed semiconductor device. There is.

If it is possible to quantitatively predict the occurrence of peeling as described above, without actually producing a molded article of a resin material,
The optimum resin material, lead frame material, lead frame surface treatment conditions such as plating, resin molding conditions, and combinations thereof can be selected.

[0133]

According to the present invention, two adhesive strengths can be obtained when residual stress is added to the stress due to a load and when residual stress is subtracted, and the stress distribution at the adhesive interface can be taken into consideration. True adhesive strength and residual stress can be separated without performing stress analysis, and universal adhesive strength that does not depend on the size and shape of the test piece can be measured with high precision and ease.

Further, according to the present invention, since the stress of the molded product obtained by the analysis and the true adhesive strength can be compared with each other, the reliability of the adhesive interface inside the molded product can be improved without actually producing the molded product. Can be estimated and evaluated.

[Brief description of drawings]

FIG. 1 is a front view showing a shape of a test piece and a load application method in an adhesive strength measuring method for a resin material which is an embodiment of the present invention.

FIG. 2 is a schematic diagram of a relationship between stresses at the start of peeling progress with respect to FIG.

FIG. 3 shows an adhesion test piece between an epoxy resin for semiconductor encapsulation and a Fe-42Ni alloy plate for a semiconductor lead frame.
It is explanatory drawing which shows the stress distribution of the peeling tip vicinity at the time of the peeling progress start analyzed by the finite element method.

FIG. 4 is an explanatory diagram showing a relationship between an apparent stress distribution and a thermal stress distribution due to only a load, and a limit stress distribution which is a true adhesive strength, for the same test piece as FIG.

FIG. 5 is an explanatory diagram showing the relationship between the apparent stress intensity factor due to only the applied load, the stress intensity factor due to only thermal stress, and the limit stress intensity factor that is the true adhesive strength for the same test piece as in FIG. 3; is there.

FIG. 6 is a front view showing a shape of a test piece and a load application method in an adhesive strength measuring method for a resin material which is another embodiment of the present invention.

FIG. 7 is a front view showing a shape of a test piece and a load application method in an adhesive strength measuring method for a resin material which is still another embodiment of the present invention.

FIG. 8 is a front view showing a shape of a test piece and a load application method in an adhesive strength measuring method for a resin material which is still another embodiment of the present invention.

FIG. 9 is a plan view of the characteristic description document of the resin material in which the measurement result of the adhesive strength by the method of the present invention is described.

FIG. 10 is a characteristic diagram showing the relationship between the true adhesive strength and the temperature of the epoxy resin for semiconductor encapsulation and the Fe-42Ni alloy plate for semiconductor lead frames.

FIG. 11 is a cross-sectional view showing the structure of a resin-encapsulated semiconductor device in which the peeling occurrence temperature at the adhesive interface is predicted using the adhesive strength according to the method of the present invention.

12 is an explanatory diagram showing a result of prediction of a peeling occurrence temperature of the resin-encapsulated semiconductor device of FIG.

[Explanation of symbols]

1 ... test piece, 2 ... resin, 3 ... adherend, 4 ... adhesive interface, 5
... Peeling portion, 6 ... Support point, 7 ... Peeling tip, 8 ... Semiconductor element, 9 ... Tab, 10 ... Lead, 11 ... Sealing resin, 12 ...
Peeling portion, 13 ... Peeling tip, P ₁ ... Load, P ₂ ... Load, τ
_r ... residual stress, τ ₁ ... shear stress due to load P ₁ , τ ₂ ... shear stress due to load P ₂ , τ _c ... limit shear stress, x ... coordinate axis parallel to adhesive interface, y ... coordinate axis perpendicular to adhesive interface , Τ _xy
... Shear stress, σ _y ... Vertical stress, K _i1 ... Stress intensity factor by load P ₁ , K _i2 ... Stress intensity factor by load P ₂ , K _ic ...
Stress intensity factor corresponding to true adhesive strength, _Kir ... Stress intensity factor corresponding to residual stress.

Continuation of the front page (56) References Japanese Patent Laid-Open No. 1-135055 (JP, A) Japanese Patent Laid-Open No. 1-280343 (JP, A) Asao Nishimura, Haise Hirose, Naotaka Tanaka, New adhesive strength measurement method for IC sealing resin , Proc. 70th Annual Conference, Japan, The Japan Society of Mechanical Engineers, September 25, 1992, Vol. B, p.363-365 (58) Fields investigated (Int.Cl. ⁷ , DB name) G01N 19/04 G01N 3/00-3/62 JISST file (JOIS)

Claims (6)

(57) [Claims] 1. An adhesive strength based on shear stress of a resin to an adherend is described together with an adherend name, and the adhesive strength is two or more including at least one resin.
For the layered construction of
Also pre-prepared at the adhesive interface between one resin and another material
Bending load in the direction perpendicular to the adhesive interface by providing peeling points
Shearing in the opposite direction by inverting and loading
A method of displaying a resin material, which is obtained by applying stress . 2. The adhesive strength of a resin to an adherend obtained by substantially excluding the influence of residual stress is expressed, and the adhesive strength is two or more including at least one resin.
At least one tree for the composition consisting of
Pre-partially peeled points at the adhesive interface between the fat and the other material
The residual stress acting on the bonding interface is
The adhesive strength is calculated by applying the stress in the direction individually.
A method of displaying a resin material, characterized by: 3. The adhesive strength based on the shear stress of the resin to the adherend is described together with the adherend name, and the adhesive strength is two or more including at least one resin.
For the layered construction of
Also pre-prepared at the adhesive interface between one resin and another material
Bending load in the direction perpendicular to the adhesive interface by providing peeling points
Shearing in the opposite direction by inverting and loading
A container for a resin material, which is obtained by applying stress . 4. The adhesive strength of a resin to an adherend obtained by substantially excluding the influence of residual stress is expressed, and the adhesive strength is two or more including at least one resin.
At least one tree for the composition consisting of
Pre-partially peeled points at the adhesive interface between the fat and the other material
The residual stress acting on the bonding interface is
The adhesive strength is calculated by applying the stress in the direction individually.
Container of resin material, characterized in that. 5. The adhesive strength based on the shear stress of the resin to the adherend is described together with the adherend name, and the adhesive strength is two or more including at least one resin.
For the layered construction of
Also pre-prepared at the adhesive interface between one resin and another material
Bending load in the direction perpendicular to the adhesive interface by providing peeling points
Shearing in the opposite direction by inverting and loading
A written document of a resin material, which is obtained by applying stress . 6. The adhesive strength of a resin to an adherend obtained by substantially excluding the influence of residual stress is expressed, and the adhesive strength is two or more including at least one resin.
At least one tree for the composition consisting of
Pre-partially peeled points at the adhesive interface between the fat and the other material
The residual stress acting on the bonding interface is
The adhesive strength is calculated by applying the stress in the direction individually.
A written document of a resin material that is characterized by