Dual solder layer for fluidic self assembly and electrical component substrate and method employing same

Abstract

A dual solder layer for fluidic self assembly, an electrical component substrate, and method employing same is described. The dual solder layer comprises a layer of a self-assembly solder disposed on a layer of a base solder which is disposed on the solder pad of an electrical component substrate. The self-assembly solder has a liquidus temperature less than a first temperature and the base solder has a solidus temperature greater than the first temperature. The self-assembly solder liquefies at the first temperature during a fluidic self assembly method to cause electrical components to adhere to the substrate. After attachment, the substrate is removed from the bath and heated so that the base solder and self-assembly solder combine to form a composite alloy which forms the final electrical solder connection between the component and the solder pad on the substrate.

Claims

1. An electrical component substrate, comprising at least one solder pad having a dual solder layer, the dual solder layer comprising a first layer of a self-assembly solder and a second layer of a base solder, the second layer of the base solder being disposed on the solder pad and the first layer of the self-assembly solder being disposed on the second layer, wherein the self-assembly solder has a liquidus temperature less than a first temperature and the base solder has a solidus temperature greater than the first temperature.

2. The electrical component substrate of claim 1, wherein the first temperature is less than about 150 C.

3. The electrical component substrate of claim 1, wherein the first temperature is less than about 100 C.

4. The electrical component substrate of claim 1, wherein a thickness of the first layer is less than a thickness of the second layer.

5. The electrical component substrate of claim 1 wherein a molar fraction of the self-assembly solder in the dual solder layer is less than about 10%.

6. The electrical component substrate of claim 1 wherein a mass ratio of the self-assembly solder to the base solder is less than 1.

7. The electrical component substrate of claim 1 wherein a mass ratio of the self-assembly solder to the base solder is between 0.5 and 1.

8. The electrical component substrate of claim 1 wherein the self-assembly solder is gallium or an alloy of gallium.

9. The electrical component substrate of claim 1 wherein the self-assembly solder and the base solder contain a same low melting point element selected from Bi, In and Sn.

10. The electrical component substrate of claim 1 wherein the self-assembly solder and the base solder contain Bi, and a mass fraction, RM, of the self-assembly solder to the base solder is determined by a formula $R_M=\frac{x_{\mathrm{Bi}-\mathrm{SA}}}{x_{\mathrm{Bi}-\mathrm{Comp}}}-1$ wherein x.sub.Bi-SA is a mass concentration of Bi in the self-assembly solder and x.sub.Bi-Comp is a mass concentration of Bi in a composite solder formed by combining the self-assembly solder and the base solder.

11. The electrical component substrate of claim 1 wherein a molar fraction of the self-assembly solder in the dual solder layer is at least 10%.

12. The electrical component substrate of claim 1 wherein a molar fraction of the self-assembly solder in the dual solder layer is at least 50%.

13. The electrical component substrate of claim 1 wherein a melting point of the base solder is above an operating temperature of an electrical component to be mounted to the electrical substrate and below a temperature at which the electrical component or substrate is damaged.

14. The electrical component substrate of claim 1 wherein the self-assembly solder contains at least one of a low melting point element selected from Ga, In, Bi, Se, Sn and Zn.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIGS. 1A-1C schematically illustrate an embodiment of a two-layer, solder-based self-assembly substrate and method according to this invention.

(2) FIGS. 2A and 2B are exemplary binary phase diagrams for self-assembly (SA) solders and base solders, respectively.

(3) FIG. 3 is a binary phase diagram for GaSn.

(4) FIG. 4 is a binary phase diagram for GaZn.

(5) FIG. 5 is a binary phase diagram for GaIn.

(6) FIG. 6 is a graphical representation of the calculated mass fraction of elements in a composite solder after reflow for Sn(0.5)-Ga(0.5).

(7) FIG. 7 is a graphical representation of the calculated mass fraction of elements in a composite solder after reflow for Sn(0.21)-Ga(0.79).

DETAILED DESCRIPTION OF THE INVENTION

(8) For a better understanding of the present invention, together with other and further objects, advantages and capabilities thereof, reference is made to the following disclosure and appended claims taken in conjunction with the above-described drawings.

(9) An embodiment of the self-assembly method employing an embodiment of an electrical component substrate in accordance with this invention is illustrated in FIGS. 1A-1C. For the sake of example, we consider SA of identical one-contact components. The two-layer solder has already been applied to the solder pads 16 of electrical component substrate 18. In FIG. 1A components 10 are placed in a liquid bath 20 whose temperature is above the melting point of the upper SA solder layer 12 but below the melting point of the base solder layer 14. The upper SA solder layer 12 is therefore liquid at the temperature of the bath whereas the base solder layer 14 which is attached to solder pad 16 on substrate 18 remains solid. Generally, the bath 20 consists of fluids such as water or ethylene glycol that do not react or damage components, solders, or substrate materials. The fluid is normally agitated to increase the probability of component attachment to solder sites on the substrate. The components 10 have their electrical contacts coated with gold or other noble metals to which the liquefied upper SA solder layer 12 will effectively wet. Preferably, the components 10 are LED dies.

(10) After some time all or most of the parts will have attached to desired solder binding sites and the SA process is terminated. The time necessary to achieve the desired attachment yield is determined by the agitation rate, the physical parameters of the liquid bath, temperature, solder surface energies, component contact characteristics, and other factors. The substrate is then cooled to allow the SA solder to solidify. In some embodiments, the SA solder is not required to solidify, provided the binding strength of the liquid SA solder is high enough to prevent components from detaching when removing the substrate from the bath. The resulting configuration for the case of solidified SA solder is shown in FIG. 1B.

(11) In a final step, the base solder layer 14 is melted so that components 10 form electrical solder connections 22 to the solder pads 16. This can be accomplished by a variety of methods known in the art, but typically would consist of passing the substrate through a reflow oven with desired time-temperature profiles. During this final reflow period both base and SA solder layers melt to form a liquid whose composition now consists of the mixture of both solders. During cooling, the solder melt will solidify forming electrical solder connections 22 comprised of a new composite alloy with the desired electrical and mechanical characteristics. This is depicted in FIG. 1C.

(12) Since the two solders form a composite alloy, the compositions and molar fractions of both base and SA solders should be chosen appropriately. Possible materials for the low melting point required for the SA solder preferably include systems with low toxicity (Hg, Pb, Sb, and Cd-free), high surface energy for reliable SA, and dissolution into common Pb-free reflow solders that would be used for the base solder. Dissolution of the SA solder into the base solder is normally possible if the molar fraction of the SA solder is not too large; less than 10% is typical. Alternatively, one could choose the molar concentrations of the SA and base solders such that they are on a eutectic; however, this requires careful monitoring of SA and base solder mass ratios on each solder pad. Therefore, the eutectic method is less desirable. In a third method, the SA solder molar fraction may have values greater than 10%, perhaps 50% or more. The solidified composite forms a new alloy, albeit different than the original alloy of the base solder. This is the preferred method, because the application of the SA solder is least constrained and can be applied by simple tinning methods in which the substrate just contacts a liquid SA solder bath.

(13) In many cases, it may be desirable to have the SA bath be at temperatures even close to room temperature. This would generally limit the SA solder to either pure elemental or metal alloys consisting primarily of Ga, In, Bi, and Sn, since many alloys of these constituents are known with melting points below even 100 C. Use of pure Ga or Ga-alloys in particular may be particularly advantageous because of gallium's low toxicity and very low melting point of 29.8 C. Ga is also soluble in other alloys. Preferably, the base solder is a solder with a melting point well above the expected operating temperature to minimize long term failure such as creep or oxidation but well below any component or substrate damage threshold.

(14) In a first embodiment, the SA solder is chosen to have a lower melting point than the base solder. No requirements are made regarding compositions such as being eutectic, binary, or other physical factors or properties. In particular, the self-assembly process occurs at a temperature T.sub.1>T.sub.SA,L, where T.sub.SA,L, is the liquidus temperature of the SA solder. Additionally, the base solder is chosen so it's solidus temperature T.sub.B,S>T.sub.1. Furthermore, during the reflow or final attachment process, the temperature for forming the final electrical solder connections, T.sub.2, is higher than the liquidus for the base solder, or T.sub.2>T.sub.B,L and will always exceed the liquidus temperature of the initial SA solder alloy. The final criterion is that the temperature T.sub.2 also be high enough to keep the entire SA and base solder solution in the liquid state; that is T.sub.2>T.sub.comp,L, where T.sub.comp,L is the liquidus temperature of the final composite alloy that forms from the liquid solution of the SA and base solders.

(15) For the sake of illustration, FIGS. 2A and 2B show two hypothetical phase diagrams for binary SA and base solder alloys, respectively, together with the above defined temperatures. The SA solder is assumed to be a binary alloy consisting of components A and B with compositions ranging from 0% to 100% of B, inclusive. Similarly, the base solder is assumed to consist of components C and D with compositions ranging from 0% to 100% of D, inclusive. In the example, the SA and base solders are assumed to have compositions xA+(1x)B and yC+(1y)D, respectively, where x and y are the respective molar fractions of A and C. In the case of the SA solder, the solid form consists of mixtures of A-rich phase or solution and B-rich phase. Note that choosing eutectic compositions for the SA and base solders are not necessarily advantageous because the final composition of the composite SA and base alloys determine the final reliability properties of the soldered components.

(16) In general, SA solder systems for this embodiment may consist of a pure metal, binary, ternary, or higher order alloys which use at least one of the following low melting point elements: Ga, In, Bi, Se, Sn, and Zn. Antimony is also a possibility in some applications, but its toxicity is greater than Ga, In, Bi, Sn, and Zn. Alkali metals generally have very low melting points, but their reactivity is highly undesirable for solder. The SA solder alloys may also include small impurities (<1%) of the elements including, but not limited to Ag, Au, Al, Cu, Ge, Ni, or Si. Such impurities are normally not desirable for the SA solder because they will increase the melting point of the SA solder, but may be either necessary to enhance final composite solder properties or dissolution behavior as the two-layer system is heated to the higher reflow temperature. Furthermore, such impurities may be unavoidable for practical solder systems.

(17) For the next embodiments of the invention, we consider SA solder compositions that have as low a melting point as possible, but will not incur undesirable properties in the final composite solder joint. The base solder is preferably a compatible lead-free solder. To reduce the number of elemental choices for the SA solder, we consider the lowest melting point elements that are usable in low toxicity solders: Ga, In, Bi, and Sn (Table 1). Note that indium generally induces creep in solder joints and should probably be avoided in most applications. It also is expensive. However, certain indium compositions have been shown to have good solder properties which are therefore useful in two layer solder self-assembly.

(18) TABLE-US-00001 TABLE 1 PRIMARY ELEMENTS FOR SELF-ASSEMBLY SOLDERS Element Melting point (T.sub.m in C.) Ga 29.8 In 156.7 Sn 232 Bi 271

(19) One important characteristic for two-layer solder self-assembly is the mass ratio R.sub.M of SA solder, M.sub.SA, to base solder, M.sub.B, where R.sub.M=M.sub.SA/M.sub.B. Generally this ratio will be much less than one because the SA solder normally contains substantial amounts of a low melting element to keep the SA solder melting point at reasonable values. On the other hand, these elements generally have smaller mass fractions in the final composite solder in order to have reasonable solder properties and higher melting points needed to minimize creep and other issues when components operate above ambient temperatures. Self-assembly however is better suited if the mass ratio R.sub.M is closer to one-half or more. This is because enough molten material must be presented to the components in the liquid bath to form adequate binding sites. Furthermore, application of the SA solder would generally be done in molten form and will strongly wet the underlying solid base solder; therefore controlled application of small amounts of SA solder is more difficult. Additionally, one would want to keep the total mass of base and SA solders to amounts similar to conventional one-layer reflow or other solder processes. Therefore, placing large amounts of base solder on the substrate to accommodate deposition of larger amounts of SA solder is not practical. With these concepts in mind, we consider examples practical systems for two-layer solder self-assembly in the next embodiments.

(20) For a second embodiment, we consider Bi binary alloys. BiSn is one of the best choices as it has a eutectic temperature of 139 C. at 57 weight % Bi. While Bi solders have some reliability tradeoffs, including brittleness and moderate toxicity, they may have applications in certain two-layer self-assembly processes. Thus one may use Bi(58)-Sn(42) eutectic solder for the SA solder of mass M.sub.SA. Note all fractional solder alloy compositions are in weight percent (wt. %). For example, Bi(58)-Sn(42) contains 58 wt. % Bi and 42 wt. % Sn. For a compatible base solder one may use compatible alloys of SnAg or SnAgCu. If one uses for example a commercially available alloy (Indium Corp.) Sn(96.5)-Ag(3.0)-Cu(0.5) (T.sub.L=220 C.), the composite solder after reflow will be close to a well known system Sn(90)-Ag(2.0)-Bi(7.5)-Cu(0.5) which has a melting point in the range of 198-212 C. This is suitable for reflow bonding and is within an acceptable temperature range for components such as LEDs.

(21) The required mass fraction, R.sub.M, of SA solder to base solder can be found by the formula:

(22) $\begin{array}{cc}{R}_M=\frac{x_{\mathrm{Bi}-\mathrm{SA}}}{x_{\mathrm{Bi}-\mathrm{Comp}}}-1,& \left(1\right)\end{array}$

(23) where x.sub.Bi-SA is the mass concentration of Bi in the SA solder and x.sub.Bi-Comp is the mass concentration in the composite solder after reflow. For the case of 7.5% Bi in the composite solder, the mass fraction of SA solder to base solder is approximately 0.15. The final composite solder composition is Sn(89.4%)-Ag(2.6)-Bi(7.5)-Cu(0.4) which is not too far from the desired composition.

(24) In general, one would make trade-offs between higher mass fractions of SA solder which may be beneficial for deposition of SA solder and self-assembly requirements, while lower mass fractions reduce Bi content and may give better composite solder properties. For example, a solder Sn(91.8)-Ag(3.4)-Bi(4.8) has shown excellent long-term electrical and thermal properties with a melting point in the range of 200-216 C. In this case one could start with a base solder of Sn(96.5)-Ag(3.5) eutectic alloy (available from Indium Corp.) which has a melting point of 221 C., which is somewhat high but acceptable for reflow. With a desired mass fraction R.sub.M=0.090, the resulting composite solder is very close to the desired Sn(92.0)-Ag(3.2)-Bi(4.8).

(25) In general, one can use a eutectic SnBi alloy for the SA solder, together with a variety of available SnAg alloys for the base solder, with Sn mass concentrations of at least 80% and preferably greater than 90% to achieve usable reflow temperatures. Additionally, variations on the eutectic SnBi alloy such as ternary alloys Sn(42)-Bi(57)-Ag(1.0) have melting temperatures (T.sub.1=140 C.) very close the binary eutectic, but may have additional desired properties. Such ternary alloys have been used in commercial products.

(26) In a third embodiment, we consider alloys with In, which may give good solder properties for the final composite solder alloy. Quaternary alloys of the form SnAgBiIn, have been investigated wherein the In mass concentration varied from 2.5-8.0%. In particular, the higher In composition Sn(88)-Ag(3.5)-Bi(0.5)-In(8) was found to have good strength and joint reliability in long-term testing, while having a reasonable solidus melting point T.sub.S=165 C. and a liquidus point of T.sub.L=206 C. Because In has a low melting point (Table 1), it is very useful for SA solder and can permit a more balanced mass ratio of SA solder to base solder than Bi-based SA solders.

(27) A good choice for an In-based SA solder is the eutectic In(50.9)-Sn(49.1) alloy which has a melting point T.sub.m=120 C., somewhat lower than the BiSn eutectic. From Equation (1) for the case of In, we find that the mass fraction R.sub.M=0.195 and the desired base solder composition is Sn(96.2)-Ag(4.2)-Bi(0.6). The liquidus point for this ternary alloy is about 235 C. which is slightly higher than desirable due to the formation of a Ag.sub.3Sn phase. Reducing the Ag concentration in the base solder to 3.5% gives a better liquidus point of about 220 C. Using the reduced Ag concentration and altering the balance of Sn in the base solder gives a final composite solder Sn(88.6)-Ag(2.9)-Bi(0.5)-In(8) which should have properties close to the solder described above. Note that more generally, a variety of InBiSn and InBi alloys have even lower melting points the InSn eutectic and may be useful for SA solders with the proper base solder composition.

(28) In a fourth embodiment, we consider Ga alloys for the SA solder because of gallium's very low melting point (Table 1) and desirable metallurgical properties when used as an additive with other solder alloys. The very low Ga melting point implies a wide range of Ga alloys will also have low melting points over a large composition range, permitting self-assembly at temperatures closer to room temperature. As discussed this is strongly desirable for two-layer solder self-assembly. Binary alloys of Ga include GaSn, GaZn, and GaIn. Phase diagrams for these alloys are shown in FIGS. 3-5, respectively.

(29) For the sake of example with Ga, we consider the simplest SA solder and base solder compositions. Referring to FIG. 3, we consider a base solder composition of Sn(1y)-Ga(y), where y is the mass fraction of Ga. If the maximum usable SA solder liquidus were assumed to be below 150 C., then the fraction of Ga, y>0.25. For more reasonable temperatures of T.sub.L=100 C. or 50 C., y=0.5 and 0.79 respectively. A reasonable base solder is the eutectic solder Sn(96.5)-Ag(3.5) with a melting point of T.sub.m=221 C., well above the liquidus temperatures for the range of Sn(1y)-Ga(y) alloys considered. FIG. 6 shows calculated mass fractions of the composite alloy as a function of R.sub.M, the SA solder to base solder mass ratio assuming an SA solder composition of Sn(0.50)-Ga(0.50). A similar calculation in shown in FIG. 7 for the lower temperature SA solder composition of Sn(0.21)-Ga(0.79).

(30) In principle, a wide range of binary, ternary, and even quaternary Ga-based SA solder alloys may exist, including Ga with major components of Sn, Zn, In, and Bi. For the base solder, lead-free alloys may be based on binary systems that include Sn such as SnAg, SnAu, SnZn, SnCu, SnBi, and SnIn. Suitable alloys may further include lesser amounts of Cu, Al, Ni, and Mg to aid solder properties. However, SA solder and base solder compositions are not limited to these combinations.

(31) While there have been shown and described what are at present considered to be the preferred embodiments of the invention, it will be apparent to those skilled in the art that various changes and modifications can be made herein without departing from the scope of the invention as defined by the appended claims.