No description available.

No description available.

The Substrateless Packaging process was developed at Loughborough University as an alternative method of manufacturing electronics with an improved end-of-life materials recovery profile. The process involves injection moulding to overmould electronic components in thermoplastic polymers. Initial prototype samples manufactured in previous work exhibited undesirable small gaps around the embedded components after solidification and which were thought to be the result of adhesion problems between the thermoplastic overmould and components. The study reported here had the aims of determining quantitatively what factors affect adhesion, and to identify which thermoplastic polymers are most suitable for the process. Following a literature survey, six engineering thermoplastics, PC, PBT, PS, ABS, PA 6 and PMMA were chosen for study as overmoulding materials, and tin as the solid adherend. The literature survey also identified the mechanisms contributing to adhesion at the metal-thermoplastic interface in insert moulding as material properties, interfacial forces between the materials, wetting at the interface, temperature of the insert (consequently temperature at the interface) and insert moulding parameters. A methodology was designed to allow investigation of all these factors, with Atomic Force Microscopy (AFM) force-distance measurements being used to measure room temperature interfacial forces, high temperature contact angle for wetting, and pull-out strength tests on overmoulded tin-coated wire for overall system adhesion. Excellent repeatability was seen in the measurements obtained with all three experimental methods. Moldflow finite element simulations of the insert moulding process were also undertaken. For the AFM measurements tin particles were adhered to the probe with the aid of a Focused Ion Beam (FIB) apparatus. PA and PMMA interatomic interactions with tin were found to be noticeably stronger than the other polymers. From consideration of the different possible contributions to the measured forces, it was concluded that the trend of interatomic interactions obtained is due to a combination of electrostatic forces, capillary forces and dispersion forces acting between the materials tested. In the high temperature contact angle measurements it was observed that the contact angles for all the materials producing drops in equilibrium reduce monotonically with rise in temperature at the interface. The work of adhesion was calculated from the contact angles for PMMA using the Young-Dupre equation and values of surface tension from the literature. It was found that it does not increase monotonically with temperature as might be expected from the contact angles. The works of adhesion at 240°C for all the materials were also calculated and it was found that the materials ranking for expected adhesive strength changed significantly from that expected from consideration of contact angle alone. In the pull out tests, except for PC, the breaking loads for the materials tested rise then fall with rise in temperature of the insert. It was observed that peak breaking load for the amorphous polymers ABS, PS and PMMA occurred for insert temperature just below Tg of the polymer, and for semi-crystalline polymers PA 6 and PBT it was just above Tg. The ranking of materials by maximum pull out strength was found to be consistent with the ranking by mechanical strength (tensile strength at yield) of the thermoplastics. The Moldflow simulations yielded the significant results that the thermoplastic melt comes in contact with the insert at relatively low pressure (less than 0.6 MPa), and that the temperature of the melt near the insert drops to the temperature of the insert almost instantaneously on contact. Therefore it was concluded that the efficacy of holding pressure on assisting wetting of the insert by the thermoplastic melt may depend on the temperature of the insert interface. The results in terms of material rankings from both the material level tests (AFM force distance experiment and wetting at high temperature) did not correspond to the mechanical strength test results. It was therefore concluded that the choice of material for thermoplastic overmould cannot be made purely based on the material interactions at interface between tin and thermoplastics in solid or melt phase. It was also concluded that the observed variation in the pull-out strengths with temperature of the insert maintained during overmoulding, must be largely due to the thermo-mechanical properties of the material at the interface. Based on the results of the study, PC, PBT and PMMA were recommended as being likely to give superior performance to the ABS which was used in early trials of the substrateless packaging process. Of these, from a process economics point of view, PBT would be the most suitable.

Due to pending legislations and market pressure, lead-free solders will replace Sn–Pb solders in 2006. Among the lead-free solders being studied, eutectic Sn–Ag, Sn–Cu and Sn–Ag–Cu are promising candidates and Sn–3.8Ag–0.7Cu could be the most appropriate replacement due to its overall balance of properties. In order to garner more understanding of lead-free solders and their application in flip-chip scale packages, the properties of lead free solders, including the wettability, intermetallic compound (IMC) growth and distribution, mechanical properties, reliability and corrosion resistance, were studied and are presented in this thesis.

In the Physics-of-Failure based approach to reliability design and assessment, which aims to relate lifetime to the identified root-cause of the potential failures, the development of effective failure mechanism models is a crucial task. The extent and rate of wire bond degradation depends on both the magnitude and duration of exposure to the loads. In the existing physics-of-failure models for wire bond interconnects, lifetime is accounted for by loading amplitude alone and is usually derived based on a regular thermal cycle of a known duration. They are not ready to predict life of arbitrary mission profiles and the effects of time at temperature on the wear-out behaviour are not addressed, leading to substantial errors for thermal cycling regimes with high peak temperatures. In this thesis, a new lifetime prediction model for wire bonds is proposed based on some phenomenological considerations, which accounts not only for the damage accumulation processes but also the damage removal phenomena during thermal exposure. The methodology discards the usual cycle-dependent modelling methodology and is instead based on a time domain representation so that it can more accurately reflect the observed temperature-time effects and related phenomena. This new time-based presentation allows estimation on the bonding interface damage condition at regular time intervals through a damage based crack propagation model which includes the effect of temperature and time dependent material properties. Meanwhile, bond degradation state is indicated in the form of crack growth and shear force reduction that are predicted by the total interface damage as a function of time. The model is calibrated and validated by the experimental results from wire material characterization tests and accelerated thermal cycling which demonstrates the advantages over the cycle-based methodologies.

The aim of this research is to understand the failure modes and mechanisms of adhesive materials used to flip-chip bond a silicon die onto a polyimide substrate. The bonding material investigated in this research is called Anisotropic Conductive Film (ACF). This is a promising interconnection material and has gained extensive interest in the electronics packaging industry. Both the experimental and finite element analysis (FEA) methods were used in order to investigate the behaviour of the ACF materials when subjected to certain manufacturing and environmental testing conditions. The manufacturing condition investigated was a subsequent solder reflow process on an ACF flip-chip bonded device. The environmental testing condition investigated was the moisture test. For the manufacturing condition, both experimental and modelling results demonstrate the impact of a subsequent reflow process on the behaviour of the ACF joint. Typical failures observed after this process were cracks at the pad/particle interface. This failure mode was more sever with a higher peak reflow temperature. This was also found using FEA where high tensile stresses were predicted in these regions. FEA modelling was also used to help identify the mechanisms leading to these failures. This is primarily due to the Coefficient of Thermal Expansion (CTE) miss-match in the materials and the elastic/plastic deformation behaviour of the conductive particle. Important design variables that can minimise these failures are the Young’s Modulus and CTE of the adhesive and the height of the hump on the die. For the environmental testing condition, an autoclave test at 121°C, 100%RH and pressure of 2atm was used. More than 85% of the ACF joints failed during the first 24 hours of testing. The failure mode observed was cracking along the interface between the adhesive and substrate and pad. A macro-micro modelling approach was used to help identify the mechanisms leading to these failures. It was found that most of the damage is caused by moisture diffusion and associated swelling. Important design variables that will help minimise this mode of failure are: Coefficient of Moisture Expansion (CME) and Young’s Modulus of the adhesive and the height of the bump on the die.