JP2006043709A - Lead-free solder alloy - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a lead-free solder alloy in which the production of voids is stably prevented by a single alloy composition. <P>SOLUTION: The lead-free solder alloy has a composition comprising, by weight, 3.4 to 4.2% Ag, 0.2 to 0.4% Cu, and the balance Sn or a composition comprising 3.4 to 4.2% Ag, 0.2 to 0.4% Cu, 0.1 to 0.8% Au, and the balance Sn. Regarding the lead-free solder alloy, two melting peaks are formed in a DSC (differential scanning calorimetry) curve, and also, provided that the total of the height of the first melting peak and the height of the second melting peak is defined as 100%, the ratio of the height of the first melting peak is 30 to 70%. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a lead-free solder alloy mainly used for manufacturing a substrate for electronic equipment.

Conventionally, Sn—Pb yarn solder alloys have been mainly used for joining electronic device substrates.
However, in recent years, lead-free solder alloys (lead-free solder) have been required due to the toxicity of lead. At present, many alloys are put into practical use as lead-free solder.
For example, Sn-Ag-Bi system, Sn-Ag-In system, Sn-Cu system, Sn-Zn-Bi system, etc., are Sn-Ag-Cu from the viewpoint of solderability, reliability, raw material cost, etc. The system is most commonly used.

By improving peripheral technologies such as solder flux and soldering method, it has become possible to mass-produce electronic device boards using lead-free solder alloys typified by Sn-Ag-Cu alloys.
However, some characteristics of solder alloys are still inferior, such as high melting point and poor wettability compared to Sn—Pb alloys that have been used conventionally.
One of the characteristics that lead-free alloys are inferior to conventional Sn-Pb alloys is that voids are often generated during reflow soldering using paste or pellets. It has been.

It is known that the bonding strength decreases due to the presence of voids in the bonded portion, and particularly when bonding heat dissipation parts such as a heat sink, heat transfer is deteriorated due to voids, which adversely affects heat dissipation.
There are two main causes of voids in reflow soldering. The first is a void resulting from poor wetting due to the cleanliness of the surface of the object to be joined and insufficient activity of the flux. Although it depends on the characteristics of the alloy itself, it can also be reduced by improving the cleanliness of the workpiece and the flux itself.
The second is a void generated when gas or flux is confined inside the solder. Theoretically, by increasing the melting time of the solder, it is possible to reduce the voids by giving a sufficient time for the gas and flux accumulated in the solder to escape sufficiently. Although the void can be reduced by lengthening the heating time at the time of melting the solder, it may not be possible depending on the heat resistance of the component or the substrate.

On the other hand, an improvement method using flux is also conceivable, but the flux plays a role of reducing and cleaning the surface oxides of the solder and the object to be joined, and the generation of gas such as H ₂ O by the reduction reaction is the principle. Is inevitable. Further, if the flux does not leave a solid content after soldering, confinement of the flux residue does not occur, but it is very difficult with the current technology.
Therefore, it is desirable to improve the characteristics of the alloy itself in order to reduce voids generated inside the solder.
There is also a technique of reducing voids by extending the half-melting time by using a paste solder mixed with powders having different solidus temperatures. (For example, see Patent Document 1)
JP 2002-113590 A

However, when a plurality of powders are used as described above, uniform characteristics may not be obtained and satisfactory characteristics may not be obtained. In addition, some very fine joints such as microbumps have several tens of solder powder particles at one location, which increases the possibility of a large variation in composition at each joint. The possibility of not being able to get is increased. Therefore, in order to suppress the void generated in the solder, it is strongly desired to achieve it with a single alloy composition.
In view of the above, an object of the present invention is to provide a lead-free solder alloy that stably prevents generation of voids with a single alloy composition.

The lead-free solder alloy of the present invention is
1) It is composed of Ag 3.4 to 4.2 wt%, Cu 0.2 to 0.4 wt%, and remaining Sn,
2) Further, it is composed of Ag 3.4 to 4.2% by weight, Cu 0.2 to 0.4% by weight, Au 0.1 to 0.8% by weight, and remaining Sn,
3) In the above 1) or 2), when two melting peaks are formed in the DSC curve and the sum of the height of the first melting peak and the height of the second melting peak is 100% The ratio of the height of the first melting peak is 30 to 70%,
4) In the above 2) or 3), Ag 3.8 to 4.2% by weight, Cu 0.2 to 0.3% by weight, Au 0.2 to 0.5% by weight, and the balance Sn.

In general, it is desirable that the soldering of the electronic device substrate is performed at a temperature as low as possible so that component destruction due to heat does not occur, and the heating time is as short as possible. In order to increase the heat resistance of the joint, it is desirable that the melting start temperature is not too low with respect to the melting peak.
In the present invention, the Sn-Ag-Cu-based lead-free material that has been used most commonly from the viewpoint of solderability, reliability, raw material cost, etc. has been sufficiently satisfied by many tests. In a solder alloy, voids generated in the solder can be suppressed without changing characteristics such as the melting temperature.

  The lead-free solder alloy of the present invention prevents the generation of voids under stable conditions by specifying the composition conditions as a single alloy of the Sn-Ag-Cu alloy that has been used stably in the past. be able to.

As described above, it is known that by increasing the melting time of the solder, it is possible to reduce the voids by giving the gas or flux accumulated in the solder to escape. The present invention reduces voids based on this principle.
Although there are several lead-free solder alloy systems currently in use, the present invention specifies the composition conditions as a single alloy of a Sn-Ag-Cu alloy that is highly reliable and has a long history of use. There is a feature in using.
Therefore, it can be used not only in paste solder but also in various shapes such as pellets. Therefore, even in a board where soldering methods are mixed, a single composition alloy can be used, and adverse effects on recycling can be reduced.

  Further, as described above, when the peak temperature is increased or the heating time is lengthened at the time of reflow, adverse effects due to heat on the components and the substrate to be joined increase. In the present invention, since it is not necessary to change the reflow profile of the current Sn-Ag-Cu-based lead-free solder, voids can be reduced without increasing the adverse effects of heat, and an Sn-Ag-Cu-based alloy is used. Therefore, it is highly reliable, and it can reduce voids without adversely affecting the rise in melting temperature and solderability even when compared with the current Sn—Ag—Cu-based lead-free solder.

FIG. 1 is a diagram showing a DSC curve of Example 1 according to Embodiment 1 of the present invention.
FIG. 2 is a diagram showing a DSC curve of Example 2 according to Embodiment 2 of the present invention.
FIG. 3 is a diagram showing a DSC curve of a comparative example.
4 is a diagram showing the relationship between the Cu content and the DSC curve in Embodiment 1, wherein (a) is Cu 0 wt%, (b) is Cu 0.1 wt%, (c) is Cu 0. 3% by weight, (d) shows the case of 0.5% by weight of Cu, and (e) shows the case of 0.7% by weight of Cu.
FIG. 5 is a diagram showing the relationship between the Cu content and the ratio of the melting peak height in the first embodiment.
FIG. 6 is a diagram showing the relationship between the Ag content and the melting rate in the first embodiment.
FIG. 7 is a diagram showing the relationship between the Au content and the melting temperature in the second embodiment, where (A) is the melting start temperature, (B) is the melting peak 0, and (C) is the melting. Peak 1, (d) indicates melting peak 2, (e) indicates peak end temperature.

Embodiment 1
The first embodiment has the composition shown in claim 1 and is an alloy close to the Sn—Ag—Cu ternary eutectic composition (generally, the eutectic composition is Sn-3.5 wt% Ag-0.7 wt%). The amount of Cu is reduced because it is said to be Cu). As a result, Cu is deficient from the Sn—Ag—Cu ternary eutectic, so that excess Sn and Ag exhibit a second melting peak as a eutectic. FIG. 1 shows a DSC curve (differential scanning calorimeter, DSC 28230 manufactured by Rigaku Corporation) in the case of a solder alloy of Example 1, Sn-4.0 wt% Ag-0.3 wt% Cu in the first embodiment. .

  Further, as apparent from FIG. 4, in the DSC curve when the amount of Cu is changed with the Ag amount being constant (3.8% by weight) in the first embodiment, (a) only the melting peak 2 is obtained without Cu. (B) As the amount of Cu increases, the height of the melting peak 1 increases, and with 0.3 wt% of Cu in (c), the height of the melting peaks 1 and 2 is almost equal to 50%, In (d) 0.5 wt% Cu, the melting peak 2 decreases, and in (e) Cu 0.7 wt% or more, only the peak 1 is obtained.

The peak height of the DSC curve represents the amount of heat per unit time required for melting, and the equal height indicates the characteristic behavior of melting in two stages. This is because, after the first stage of melting is completed, if the heat quantity substantially equal to that is not applied, the second stage cannot start melting, a melting time lag occurs, and the entire melting time becomes longer.
On the other hand, when the difference in height is large, the smaller peak is buried in the larger peak, and the behavior is such that the larger peak is melted at one time, so the void reduction effect cannot be obtained.
In addition, the numerical value shown in FIG. 4 represents this ratio of each peak height, when the sum total of two melting peak heights of a DSC curve is 100%.

FIG. 3 shows a DSC curve in the case of a conventional S-3.0Ag-0.5Cu alloy as a comparative example. In this case, one melting peak of Example 1 of Embodiment 1 is shown. The form is shown.
On the other hand, in the DSC curve according to the first embodiment, the melting is divided into two distinct melting peaks in a certain range of Cu%, so that a time lag occurs in melting. Therefore, the half melting time can be made longer than that of the conventional Sn-Ag-Cu lead-free solder alloy in which only one peak of melting is seen. It can be seen that the melting start temperature and the melting end temperature of both are substantially equal.
Therefore, it can be seen that the solder alloy in Embodiment 1 does not need to change the reflow profile as compared with the current solder alloy.

However, as apparent from FIG. 4, since the amount of the Sn—Ag—Cu eutectic that starts melting first is too large or too small, it appears to be melted at one time. The upper melting time cannot be increased.
Therefore, in the present invention, the optimum amount of Cu is specified such that Sn—Ag—Cu eutectic and Sn—Ag eutectic are almost halved.

FIG. 5 shows the relationship between the Cu content and the ratio of the melting peak height in the first embodiment. If the ratio of the peak height is in the range of about 30 to 70%, the melting time is lengthened. It can be seen that the optimum amount of Cu is 0.2 to 0.4 wt%.
In particular, at 0.3% by weight of Cu, the peak height ratio is about 50%, and the most desirable effect is obtained.

On the other hand, the examination examination about the appropriate range of Ag was carried out. Focusing on the fact that the valley of the peak is about 220 ° C. from the DSC curve of FIG. 1, the peak area of 220 ° C. or less with respect to the total peak area, that is, the ratio of the first peak area to the whole (hereinafter referred to as the melting rate). ) For each composition.
The result is shown in FIG. Since the latent heat of fusion depending on the composition is not greatly different, the larger the melting rate, the more parts are melted, and the onset of melting is accelerated in appearance. From FIG. 6, it can be seen that the melting rate is higher in the range of 3.4 to 4.2% by weight of Ag compared to before and after that. Therefore, it was found that a larger portion melts at the melting peak of the first stage in the composition range of Ag.

In general, addition of Cu not only lowers the melting temperature but also has an effect of improving solderability such as wettability. The solderability is slightly reduced by the amount of Cu, but it is not a problem for practical use. However, in order to obtain better solderability, it is desirable to contain more Ag that improves the solderability as well as Cu within a good composition range.
Therefore, the Ag content is set to 3.4 to 4.2% by weight, and in order to obtain better solderability, it has been found that 3.8 to 4.2% by weight is the best.

Embodiment 2
FIG. 2 is a diagram showing a DSC curve of Example 2 according to Embodiment 2 of the present invention.
FIG. 7 shows the results of measuring the content of added Au and the temperature of each melting peak in the second embodiment.
In the second embodiment, by adding 0.1 to 0.8% by weight of Au to the composition according to claim 1, the melting start temperature can be lowered and solder wetting is further improved. be able to. In the case of adding Au, there is no risk that a low melting point phase of 150 ° C. or lower such as Bi or In is generated, so that there is no significant adverse effect on reliability such as heat cycle.

  The present invention utilizes the principle that voids can be reduced by increasing the already known melting time. And since it uses a Sn-Ag-Cu alloy, it is highly reliable, and also has an adverse effect on the rise in melting temperature and solderability compared to current Sn-Ag-Cu lead-free solder alloys. The void reduction can be achieved without any problems.

Conventionally, there is an invention of a solder alloy using two melting peaks (for example, Japanese Patent No. 2682326), but these are for the purpose of suppressing chip standing called Manhattan phenomenon. Since the present invention does not satisfy the requirements (such as peak size and ratio) required in the above patent, the effect of suppressing chip standing cannot be obtained. The present invention has a different purpose from that for suppressing chip standing, and the alloy according to the present invention can use the same reflow profile as the current Sn-Ag-Cu alloy. This is advantageous for alloys with two melting peaks.
As for the optimum amount of Ag, it was found by experiments that the apparent onset of melting was early in the range of Ag 3.4 to 4.2% by weight. By starting the melting of as many solders as possible as soon as possible, the rise of the start of wetting to the workpiece is improved, and as a result, good wettability is obtained. This phenomenon can be explained by comparing the area ratio of the DSC curve, that is, the amount of heat corresponding to the latent heat of fusion.

Example 1 in the first embodiment is a solder paste having the following composition.
Solder alloy powder:
Composition Ag 4.0 wt%, Cu 0.3 wt%, remaining Sn
Average particle size 25-45 μm
Flux: Flux for resin paste Kneading ratio: Solder alloy powder / flux = 89% / 11%

Example 2 in the second embodiment is a solder paste having the following composition.
Solder alloy powder:
Composition Ag 4.0 wt%, Cu 0.3 wt%, Au 0.2 wt%, remaining Sn
Average particle size 25-45 μm
Flux: Flux for resin paste Kneading ratio: Solder alloy powder / flux = 89% / 11%

The melting start temperatures, melting peak 1 temperatures, melting peak 2 temperatures, and peak end temperatures of the above-mentioned Examples 1 and 2 and Comparative Examples were compared. The composition of the solder paste of the comparative example is as follows.
Solder alloy powder:
Composition Ag 3.0% by weight, Cu 0.5% by weight, remaining Sn
Average particle size 25-45 μm
Flux: Flux for resin paste Kneading ratio: Solder alloy powder / flux = 89% / 11%
Each temperature comparison result is shown in Table 1.

  The values of Examples 1 and 2 in Table 1 correspond to the values in the DSC curve of FIGS. 1 and 2, and the values of Comparative Examples correspond to the values in the DSC curve of FIG. 3. In comparison with the comparative example in which only the melting peak 1 is clearly present, the melting peak 2 clearly appears in Examples 1 and 2. The melting start temperature and peak end temperature in Example 1 are almost the same as those in the conventional comparative example.

Example 2 is an alloy obtained by adding Au to Example 1, Sn-4.0 wt% -0.3 wt% Cu-0.2 wt% Au, as can be seen from the DSC curve in FIG. By adding Au, the height of the melting peak 1 is lowered and the melting start temperature is relatively lowered, but the change of the melting peak 2 is relatively small.
This is because a part of the melting peak 1 which is the Sn—Ag—Cu ternary eutectic with the addition of Au becomes the Sn—Ag—Au—Cu quaternary eutectic, so that the peak height is distributed. Because. On the other hand, the melting peak 2 is an Sn—Ag binary eutectic, so that it is hardly affected by the addition of Au and the peak height is maintained. In this way, the addition of Au behaves differently from the case of the Sn-Ag-Cu-based ternary composition alone, but the effect of lengthening the melting time is obtained by maintaining the height of the melting peak 2 and reducing voids. can do.

Generally, Bi or In is often used as an element that lowers the melting temperature of lead-free solder, but these elements are known to form a eutectic with Sn. That is, in the case of Sn-Bi eutectic, the melting temperature is very low at 139 ° C., and in the case of Sn-ln eutectic, 120 ° C. (both from ASM: Binary Alloy Phase Diagrams), and it is theoretical when Ag and Cu are added. Further, the melting temperature is further lowered. Even if these elements are added, even if the melting peak of the low melting point phase does not appear in the thermal analysis, the subsequent segregation may partially lower the melting temperature, which may affect reliability such as heat cycle. .
On the other hand, the low melting point phase observed in the case of addition of Au of the present invention is Sn-Ag-Au ternary eutectic at 206 ° C. (from ASM: Handbook of Thermal AJIoy Phase Diaqams), Sn-Ag-Au— It is confirmed that the Cu-based quaternary crystal is 204 ° C. (according to the inventor's experiment), and does not have a melting temperature of 200 ° C. or less due to segregation of the solder itself. Therefore, even if the melting start temperature is lowered and solder wetting is improved, the reliability such as heat cycle is not adversely affected.

  FIG. 7 shows changes in the melting temperature when Au is added. From this, when the added amount of Au exceeds 1.0% by weight, a peak that seems to be a Sn—Ag—Au—Cu quaternary eutectic (referred to as melting peak 0 in the figure) appears clearly at 204 ° C. This indicates that a low melting point phase appears due to solidification segregation from the initial state. This low melting point phase is at a higher temperature than Bi or In, but considering the reliability such as heat cycle, it is necessary to be 0.8 wt% or less of Au at which the low melting point phase does not appear from the initial state. In addition, since the melting start temperature extremely decreases when Au is 0.5% by weight or more, it is desirable that the added amount of Au is less than this in consideration of heat resistance and reliability. On the other hand, the lower limit value is 0.1% by weight or more of Au, and the effect of improving wetting is obtained, but 0.2% by weight or more is desirable to obtain a remarkable effect.

Next, the length of the half-melting time, which is the principle of the void reduction effect, was verified by Examples 1 and 2 and the comparative example.
The solder pastes of Examples 1 and 2 and the comparative example are printed on a 30 × 30 × 0.3 mm copper plate with a metal mask of φ6.5 × 0.2 mm. This was observed using a high temperature observation apparatus (SMT-Scope SA5000 manufactured by Sanyo Seiko Co., Ltd.), and the time from the start to the end of melting was measured. The heating conditions at this time are as follows: preheating is 200 ° C. × 60 seconds, and the heating rate of main heating is 10 ° C./min.
Table 2 shows the comparison results of the half melting times according to Examples 1 and 2 and the comparative example.

Furthermore, the generation of voids in the actual reflow process was verified.
The solder pastes of Examples 1 and 2 and the comparative example are printed on a copper land printed circuit board using a metal mask. This is placed on a hot plate at 260 ° C. for 30 seconds and heated, and then placed on a metal plate or the like for cooling. The voids in the solder portion of the reflow soldered substrate were observed with an X-ray TV inspection device (SVJ-2000 manufactured by Softex), and the amount of void generation was compared.
Table 2 also shows the verification results (void evaluation) of the void generation amount according to Examples 1 and 2 and the comparative example.

As is apparent from Table 2, even under the same heating conditions, it was found that the half melting time was longer by about 20% in Examples 1 and 2 than in the comparative example.
Further, Examples 1 and 2 yielded good results with less generation of voids as compared with the comparative example.

Further, the applicant's separate experiment confirmed that the composition of Example 2 was better wetted than the composition of Example 1.
Thus, this was verified by a spread rate comparison test by a spread test.
The test method conformed to JISZ3197, and the reflow temperature was 260 ° C.
Table 3 shows the evaluation of the spreading ratio of Examples 1 and 2 and the comparative example.

As shown in Table 3, the spreading rate was found to be favorable in the order of Example 2> Example 1> Comparative Example.
Therefore, it was confirmed that the solder wetting is also good along this order.

As described above, according to the present invention, the single alloy composition reduces the adverse effect on recycling, and also the speed of the onset of melting and the solderability of the alloy itself are appropriate, and the reduction of voids is achieved.
In addition, Example 1 and Example 2 are demonstrated by each 1 Example of the aspect 1 and 2 of embodiment of this invention, and if it is in a claim, it will be limited to this composition range. It is not a thing.

  In the present invention, void reduction is achieved with a single alloy composition. Therefore, not only paste solder but also various shapes such as pellets can be used.

It is a diagram which shows the DSC curve of Example 1 in Embodiment 1 of this invention. It is a diagram which shows the DSC curve of Example 2 in Embodiment 2 of this invention. It is a diagram which shows the DSC curve of a comparative example. It is a diagram which shows the relationship between Cu content and DSC curve in this Embodiment 1, (a) is Cu0 weight%, (b) is Cu0.1 weight%, (c) is Cu0.3 weight%, (D) shows the case of Cu 0.5 wt%, (e) shows the case of Cu 0.1 wt%. It is a diagram which shows the relationship of the ratio of Cu content and the fusion peak height in this Embodiment 1. FIG. It is a diagram which shows the relationship between Ag content in this Embodiment 1, and a melting rate. It is a diagram which shows the relationship between Au content and melting | fusing temperature in this Embodiment 2, (A) is melting start temperature, (B) is melting peak 0, (C) is melting peak 1, ( D) shows melting peak 2, and (e) shows peak end temperature.

Claims (4)

  A lead-free solder alloy comprising Ag 3.4 to 4.2% by weight, Cu 0.2 to 0.4% by weight, and remaining Sn.   A lead-free solder alloy comprising Ag 3.4 to 4.2 wt%, Cu 0.2 to 0.4 wt%, Au 0.1 to 0.8 wt%, and the balance Sn.   Two melting peaks are formed in the DSC curve, and the height of the first peak is the first melting peak when the sum of the heights of the first melting peak and the second melting peak is 100%. The lead-free solder alloy according to claim 1 or 2, wherein a height ratio is 30 to 70%. The lead-free solder according to claim 2 or 3, wherein Ag is 3.8 to 4.2% by weight, Cu is 0.2 to 0.4% by weight, Au is 0.2 to 0.5% by weight, and the balance is Sn. alloy.

Images