Research On Interfacial Reaction Of Lead-free Solders In Electronic Packaging

Along with the development towards lead-free interconnection materials inmicro-electronic packaging industry and the progress towards miniaturization andmulti-functionalization for electronic products, the reliability issues of the interconnectionsolder joints as service have aroused the wide attention. However, the interfacial reactionbetween lead-free solder and substrate （including liquid-solid and solid-solid reaction） is themost key issue for the reliability of lead-free solder joint, so, it is necessary to research theinterfacial reaction mechanism for lead-free solder joints.Firstly, the interfacial reaction between Sn-9Zn solder adding Cu-particle and Cusubstrate was researched in this work. The experimental results show that the growth rate ofthe interfacial IMC （Cu₅Zn₈） in the solder joints can be fully suppressed by addingCu-particle in the solder, meanwhile, Zn phase inside the solder joint is in situ transformedinto Cu-Zn intermetallic compounds due to its reaction with Cu-particle, resulting in Zn phasedecreasing. Since the growth rate of the interfacial IMC of the solder/Cu-substrate isobviously lower than the solder/Cu-particle, it is concluded that Cu-particle is able to“capture” the more Zn atom than Cu-substrate. This can be explained as that the quantities ofthe liquid solder reacting with Cu-particle are the more than Cu-substrate in the now reactionsystem by means of Dybkov’s interfacial reaction theory.Secondly, the effect of adding trace alloy elements Ti and Ni in Sn base solder on theinterfacial reaction was explored. The results indicate that the growth rate of the interfacialIMC between Sn-0.7Cu solder and Cu substrate can be decreased in some extent, and it is alsoobserved that the IMC grain size increases when0.008wt%Ti is added in the solder as alloyelement. However, since Ti element is not detected in the interfacial IMC, this concluded thatthere is not any reaction between Ti and other constituents forming the IMC. So, themechanism-controlled depressing the IMC growth rate may be considered that the boundarydiffusion of the constituents Cu and Sn is restrained due to decreasing the nucleation of theinterfacial IMC as adding Ti in the solder, resulting in the grain size increasing, meanwhile,the boundary area decreasing. When0.05wt%Ni is added in Sn_0.8Ag_0.5Cu₂Bi solder as alloy element, the interfacial reaction IMC between the solder and Cu substrate is （Cu,Ni）₆Sn₅assoldering, and Ni content in the IMC is about3-4at%by EDS. Since the （Cu,Ni）₆Sn₅configuration is more complex than Cu₆Sn₅, its growth rate is depressed as isothermal aging.The （Cu,Ni）₆Sn₅growth activation energy calculated by experimental data is about111.62kJ/mol, and it is more larger than Cu₆Sn₅（about75.03kJ/mol）.Additionally, the interfacial reactions between Sn_0.3Ag_0.7Cu solder and dissimilar padsfinish （including OSP, HASL, Electrolytic Ni/Au and ENIG） were comparatively investigated.It is presented that scallop-shaped Cu₆Sn₅are formed between the solder and OSP as well asHASL pads finish as soldering, however, the interfacial IMC with Electrolytic Ni/Au andENIG pads finish are composed of layer-shaped and dollop-shaped （Cu,Ni）₆Sn₅as well asneedle-shaped （Ni,Cu）₃Sn₄. During isothermal aging, the needle-shaped （Ni,Cu）₃Sn₄gradually reduce till completely vanish and the thickness of layer-shaped （Cu,Ni）₆Sn₅,simultaneously, increases with aging time increasing. It is concluded that Cu atom in thesubstrate participates in the interfacial reaction and its diffusive rate through Ni layerdominates the IMC growth rate, therefore, the needle-shaped （Ni,Cu）₃Sn₄more slowly reduceas the Ni layer is more thick.Finally, the growth kinetics of the interfacial IMC of Sn_3.0Ag_0.5Cu/Cu solder joints wasalso explored under stressing electric current. It is indicated that the evolution on theinterfacial IMC between the solder and Cu substrate presents an obvious polarity effect atcathode and anode. The growth rate of the interfacial IMC is strengthened at anode andinhibited at cathode, and its growth kinetics at anode accords with parabolic relationship withstressing time. Although the dissolution rate of the interfacial IMC at cathode also isaccordance with parabolic rule, the IMC thickness exist a threshold value. When the IMCthickness is larger than the threshold value （early stage）, the IMC dissolution followsparabolic rule, however, as the IMC thickness is equal to the threshold value, it is hardlychanged. This concludes that the driven forces acted on the diffusive atoms at cathode takebalance, namely, electronic wind force is equal to chemical potential energy.