eMail the Editor
Lead Free Solders for Surface Mount Technology Applications (Part 2)
Surface TensionThe interfacial forces are among the most critical factors in soldering. These forces between the molten solder, flux and substrate influence the degree of wetting which in turn determines the formation of proper solder joints. Interfacial forces are the interactions of the surface tension of the molten solder, the flux and the substrate material during soldering, and, as such, surface tension is one of the critical physical properties of solder that determines its wetting behavior.
In addition to wetting, the surface tension of the molten solder plays a number of critical roles in SMT assembly. Capillary flow for plated-throughhole (PTH) soldering, self alignment of surface mounted devices and the capability to keep devices from falling off during second reflow are all functions of the surface tension of the molten solder.
ThermodynamicsThe surface tension of a liquid is a thermodynamic quantity and is defined as the amount of work needed to isothermally enlarge the liquid surface area.35 Seldom is a thermal equilibrium condition reached in actual SMT soldering, because the soldering operation is completed before equilibrium temperate is reached. Furthermore, dissolution of the substrate in the molten solder, the oxidation of the flux, the soldering environment, etc., all affect the surface tension of solder.
Studies have shown that the surface tension value of solder varies with temperature,36 flux composition37 and the extent of solder substrate interactions.38 For these reasons the surface tensions of solder and most other liquid metals are not precisely known.39
There is little surface tension data available for lead-free solders. Vincent, et al. 40 measured the surface tension of a range of binary Pb-free alloys in both air and nitrogen with <20 ppm O2, at 50°C over their liquidus temperature. (The data are shown in Table 3.)
Generally, surface tension values tend to be lower in air than in an inert atmosphere since oxidation lowers the free energy of the liquid surface. The data for eutectic Sn-Zn and Sn-Cu do not fit this trend. Furthermore, the surface tension of a liquid decreases linearly with increasing temperature and continues to decrease as the temperature approaches the boiling point. Further work is needed in this area.
Table 3. Measured Values of Surface Tension for Binary Alloys
Table 4. Physical Properties of Pb-free Solders40
Coefficient of Thermal ExpansionCoefficients of thermal expansion (CTE) and density data of Pb-free solders are provided in Table 4. The CTE mismatch between the substrate, the device and solder is one of the critical factors that determines the fatigue life of a solder joint. It is the CTE of the solder that determines the fatigue life of a solder joint. It is the CTE of solder that determines the strain and stress distribution in a solder joint during thermal cycling. Having a Pb-free solder with a CTE that closely matches the CTYE of Cu or Alloy 42 device leads and substrate metallization is extremely desirable. There are many Pb-free solders that provide a better CTE match than eutectic tin-lead solder.
ResistivityResistivity is one of the fundamental electrical properties of solder. For most electronics applications, the resistivity of solder is relatively low and its effect on the overall functionality of the circuit is insignificant. Resistivity varies with temperature, composition and microstructure of a given alloy. Metals that are pure with ordered crystal structure of a large grain size have low resistivity at low temperature. Generally, alloys have higher resistance that typically peaks when the elements present are in roughly equal fractions. Because of the dependence of resistivity on microstructure, grain size, dislocation density, etc., measured values differ significantly among both pure metals and alloys of the same composition.40 Room temperature resistivity values for common solders and packaging materials used for microelectronics applications are shown in Table 5.
Table 5. RoomTemperature Resistivity Values for Some Pb-Solders Microelectronics Packaging Materials.41
CorrosionDuring processing and service life, it is highly likely that solder joints will be exposed to moisture, ionic species and other contaminants that cause and promote corrosion. Therefore, the ability of lead-free solder to resist corrosion is a desirable property. However, there is virtually no corrosion test data available for most of the lead-free solders. Corrosion behavior of 52In-48Sn soldered on Au substrate in the presence of 15 to 25 ppb. Cl2 at 85°C and 85% relative humidity was studied by Abtew.47
Electromotive ForceDifferences in the electromotive force (EMF) between the phases present in the alloy are frequently a good indication of the corrosion potential. If the EMF between the phases present in the solder alloy is large corrosion is likely to take place because the electrochemical coupling in the presence of moisture is high. Data that indicates differences in EMF between the different phases of the Pb-free alloys was not available. In Table 6, the difference in EMF between the constituent elements of some Pb-free solders is presented:48
Mechanical PropertiesThe mechanical properties of solder joints represent some of the most critical factors in SMT soldering. The application of mechanical forces to a solid body causes the body to deform and may be even to fracture. Of special importance are the stresses and strains that are used to characterize the behavior of material under different types of mechanical loading.
The mechanical property of a solder joint defines the response of solder joints to imposed strains and stresses. The properties of major concern for solder in SMT applications are shear strength, ultimate tensile strength (UTS), ductility, creep and fatigue resistance.42 Glazer divided the mode of imposition of stresses and strains into three broad categories: time independent monotonic deformation, time-dependent monotonic deformation and cyclic deformation.40
Time-Independent Monotonic DeformationThe types of deformation in this category are tensile and shear tests. When solid materials are subjected to small stresses they usually respond in an elastic fashion, i.e., the strain produced by the stress is reversible and the magnitude of the strain is proportional to the magnitude of the stress. This reversible deformation, where stress and strain is held constant, is called elastic deformation. With increased stress, the material starts to undergo plastic deformation. Once plastic defamation takes place, the material is deformed permanently and will not recover its original shape when the stress is removed.
Elastic ModulusWhen only elastic deformation exists, the strain is proportional to the applied stress. The ratio of stress to strain is the elastic modulus. The strength of a material is directly related to the elastic modulus43 and the elastic modulus, in turn, is directly related to the forces between atoms in the crystal lattice. In pure metals or metals with small compositional changes, elastic modulus is invariant to microstructural changes. For alloys such as eutectic solders, which contain a mixture of phases in significant proportions, however, the elastic modulus may be dependent on microstructure, because each component of the alloy may have different moduli. The elastic modulus is also an important parameter in finite element modeling of solder joint stresses and strains. The room temperature elastic modulus values40 for common soldering elements and some of their alloys are presented in Table 7, along with yield strength and shear strength40 data.
Time-Dependent Monotonic DeformationThis type of deformation is commonly referred to as "creep," a measure of the time required for a material to fail when it is under a constant load at a constant temperature.44 Creep involves deformation mechanisms, such as grain boundary sliding, vacancy diffusion, etc., which require a thermally-driven diffusion process. Therefore, creep deformation becomes critical only when the temperature exceeds half the absolute melting temperature of the material.44 For most soldering alloys, room temperature is well above half their absolute melting temperature. Consequently, for solders in SMT, creep is considered the most important deformation mechanism.45
d/dt = Atn exp(-H/RT) where , is the shear strain rate, t is the shear stress, and the stress exponent (n) and activation energy (H) are specific to the dominant creep mechanism. The constant A is microstructure dependent, R is the universal gas constant and T is the temperature in degrees Kelvin. For Sn-Pb solder, the value of n varies from 2 to 7 depending on the microstructure and the dominating creep mechanism. Similarly, the activation energy is related to the rate controlling process for secondary creep40 and gives insight into the creep mechanism of the material.
In eutectic Pb-Sn solder, the rate controlling mechanism at strain rates below 10-5/sec is lattice self-diffusion50 which appears to be the case for Sn-Bi also. Primary, secondary and tertiary creep were observed for 42Sn58Bi51. Mei and Morris observed that at 65°C solder joints of eutectic Sn-Bi are more creep resistant than eutectic Pb-Sn solder joints formed under the same conditions.51 The total strain-to-failure, however, was only 20-25%, which is less than what is observed for eutectic Sn-Pb.
Tests conducted by Freer and Morris48,49 to evaluate the creep behavior of eutectic In-Sn solder indicated rapid and extensive deformation leading to early failure. Primary creep was not observed; however, a steady-state creep region and a tertiary creep region in which extensive deformation took place was encountered. Furthermore, no microstructural changes like coarsening and recrystallization similar to Pb-Sn, were observed. Test results from Darveaux and Banerji50 indicated that the Sn-3.5Ag alloy exhibited considerably more strain before failure than eutectic Sn-Pb under their test conditions. The test results also suggested the acceleration factor in a thermal cycling test may be greater for Sn-3.5Ag than for eutectic Sn-Pb under the same conditions.
Cyclic Deformation or FatigueFatigue, a measure of resistance to cyclic loading, can be isothermal or "thermal." Isothermal fatigue is where imposed cyclic displacement occurs at a constant temperature. Thermal fatigue, on the other hand, is a condition where cyclic displacement occurs due to a change in temperature, because of the joining of two materials with dissimilar thermal coefficients of expansion.
Fatigue in solder joints leads to crack initiation and crack propagation; the fatigue life of a solder joint is determined by the number of stress cycles it endures before a crack is initiated and propagates. Even when the cyclic stress is well below the yield stress of the material, fatigue failure can occur due to defects and irregularities in the microstructure that may serve as crack initiation sites.
Failure in solder alloys involves both fatigue and creep. For eutectic tin-lead solder, the failure mode in creep and in fatigue appears to be the same. However, the relationship between thermal fatigue life and creep data is not well understood.46
There is a scarcity of fatigue-resistance data for most of the Pb-free alloys. The little data that exists lacks accurate characterization of the test conditions, initial microstructure or the failure mechanisms. According to Lea,52 eutectic Sn-Pb has the lowest fatigue resistance. The fatigue resistance of some Pb-free solders ranked in increasing order were: 64Sn-36In, 42Sn-58Bi, 50Sn-50In, 99.25Sn-0.75Cu, 100Sn, 96Sn-4Ag and 95Sn-5Sb.
The test that mimics surface mount applications relatively well was conducted by Strauss and Smernos.52 Their findings indicated that eutectic Sn-Bi performed better than eutectic Sn-Pb in thermal cycling of through hole components from -40°C to 70°C and 16 min/cycle. Marshall and Walter53 conducted thermal cycling tests on a Cu pin soldered into a 0.060" diameter plated-through hole encased in a zinc housing filled with polypropylene. The stress on the joint was created by the expansion of the polypropylene when exposed to high temperature. The test conditions were -30°C to 100°C with 60-minute dwells and a ramp rate of 6.5°C/min. The test results indicated that Ag-containing solders are superior to eutectic Sn-Pb and far superior to In-Sn solders.
Table 7. Elastic Modulus Values at Room Temperature for Elements Commonly Found in Solder and its Alloys.
ConclusionA relatively large number of Pb-free solders have been proposed thus far, and some are available from manufacturers. These solders span the spectrum of binary, ternary and quartenary alloys. A common feature of all Pb-free solder alloys is the lack of uniformity in the engineering data available, which is essential for the engineer to be able to design reliable solder joints.
The Sn-58Bi, In-48Sn and Sn-3.5Ag lead free solders appear to be, by far, the best investigated systems. In most of the studies, the data was acquired from tests conducted on bulk solders or under conditions that have little resemblance to actual SMT applications. Lack of information on initial microstructure of test specimen, stress and strain conditions, failure mechanisms, etc., limit the applicability of the findings.
Based on the reviewed data, eutectic Sn-Bi appears to have properties that closely resemble eutectic Sn-Pb. The relatively high melting point of eutectic Sn-Ag, may be a problem. However, it must be seriously considered for applications where the solder joints may encounter large stresses and strains.
If Pb-free solders are indeed to replace eutectic Sn-Pb alloys, a systematic investigation of the fundamental engineering properties, and performance in "real-life" conditions is necessary.
Chip Scale Review o 7291 Coronado Drive, Suite 8 o San Jose, CA 95129 o Email: email@example.com
|© 1998 ChipScale REVIEW|