Au-Sn eutectic solder is widely used in optoelectronic packaging. Electroplating of Au-Sn alloys is attracting increasing research interest because of its lower cost and potential to minimize voiding in bonding layers. The challenge in developing a non-cyanide Au-Sn electroplating bath lies in bath stability and deposit composition control. In this paper a relatively stable electroplating bath is developed based on a chloride solution using citrate as a buffering agent. Direct current plating and pulse current plating are performed and compared. The effects of Au and Sn concentrations in the bath, current density, and ON time on deposit composition and microstructures are also studied.

Solder paste performance related properties such as stencil print quality, viscosity, thixotropy, and slump resistance are of great importance in the pre-reflow stages of the electronics assembly process. This paper focuses on the study of the rheological behavior of solder paste and its correlation to process performance during the various steps of the Alternative Assembly and Reflow Technology (AART) process. This technology aims to integrate the Printed Circuit board (PCB) assembly process for both through hole and surface mount components. Four solder pastes were considered in this study and their rheological characteristics were identified through flow and oscillation tests. The oscillation tests provided the linear viscoelastic characteristics of solder paste whereas the flow tests revealed information on the yield stress as well as the degree of shear-thinning and thixotropy of the solder paste. Pastes with high elastic properties and yield stress exhibited good hot slump resistance. Furthermore, extensive shear thinning of the paste facilitated the filling of the Plated-Through-Hole (PTH) sites. Recommendations are provided for tailoring the properties of a solder paste to meet the needs of the AART process.

Features and recent developments of various photomechanics methods are reviewed for thermal deformation analyses of electronic packages. Applications to diverse problems are illustrated to demonstrate wide applicability of the methods. The whole-field displacement information, with various sensitivity and resolution scales, is ideally suited for the deformation study of a broad range of problems in deformation analyses of electronic packaging. The methods are mature and they can be practiced routinely. A more extensive range of applications are anticipated.

Integration of core electronic functions is progressing at a remarkable rate. The high-end computers that we are using now may be available on a single die by the turn of the century. The effect of this, and other advances on semiconductor assembly and packaging has generated the need for integrated component-level electrical thermal, and mechanical models and simulation methods, as described in the Semiconductor Industry Association national Technology Roadmap for Semiconductors. Although corporate and university research and development in the semiconductor area is intense (there are eight major university-based packaging centers around the world, for example), design/development methodologies that consider the combined and interactive effects of mechanical, thermal and chemical conditions on the reliability and durability of integrated components have been slow to develop. However, such methodologies have been developed for applications of complex composite systems in various other technologies where the operating conditions are similar. The present paper will discuss those developments and the possibility of applying that technology to microelectronic systems.

The residual stress evolution in an epoxy during isothermal curing process has been determined experimentally. It was found that for a certain curing temperature range in which the characteristic time for the molecular motions leading to volume recovery is shorter than the time scale of the experiment, an incremental elastic constitutive equation can be used to describe the mechanical response of the epoxy. It was also found that appreciable residual stresses are developed in a threedimensionally constrained epoxy resin system within the rubbery state.

Metal-polymer bilayers and multilayers are being increasingly used in the microelectronics industry. In many cases, the reliability of the device or the package is determined by the mechanical behavior of these composite layers. Finite element analyses are being widely used to model these systems. An accurate experimental understanding of the mechanical behavior of these systems would further enhance the usefulness of such analyses. The present study focuses on the mechanical behavior of composite bilayers consisting of polyimide and sputtered Ti and Ti alloy thin films. Tensile tests and stretch-deformation tests have been performed on bilayers consisting of a polyimide substrate sputter-coated individually with a 50 nm and a 500 nm Ti layer, a 500 nm Ti-Cu layer, and a 500 nm Ti-Ni layer. Tensile moduli, stresses for 0.2%, 2% and 15% offset strains, and strains to failure have been obtained and compared. These values are discussed with reference to the known mechanical properties of the individual layers in the composite material.

Epoxy based resins have been widely used as encapsulation materials in microelectronics components due to their good mechanical, thermal and electrical properties, outstanding adhesion to various substrates and the ability to be processed under a variety of conditions. These epoxy resins have a high glass transition temperature but low toughness and damage tolerance. As electronic devices continue to be miniaturized and the scale of integration increases, the encapsulation of these chips leads to mechanical and thermal stress related failures. Mechanical stresses could result from the mismatch in thermal expansion rates between the polymeric encapsulant, the silicon chip and the metal leadframe. The accommodation of these stresses may lead to cracking of the encapsulant or delamination. In this study, poly(oxazolidone) polymer was prepared by reacting 4,4′-Diphenyl-methanediisocyanate (MDI) with diglycidyl ether of bisphenol A (DGEBA) at elevated temperature. The mechanical properties of this polymer were investigated both as a stand-alone material as well as a bi-material sandwich with a brittle substrate. The degree of cracking in the substrate is shown to be a function of the thickness of the polymer layer. This result is analyzed based on the variation in the position of the neutral axis in this composite.

The continuous miniaturization of electronic devices places an ever-increasing importance on the thermal management of electronic systems and its subcomponents. The increased power densities and heat generation, due to the miniaturization of the device line features, may lead to higher operating temperatures and greater warpage between the silicon device and its organic carrier. The higher operating temperature may result from the degradation of the overall thermal performance. These additive effects will also lead to an increasing number of thermally induced failures, which will be further magnified when future microelectronic packaging incorporates flip-chip technology.

The higher operating temperatures within microelectronic systems result from inadequate dissipation of the heat generated, while the warpage effect is caused by the mismatch between the thermal coefficients of expansion (ICE) induced by thermal stresses. Often these high temperatures result from the thermal resistance between subcomponents, such as between the contacting surfaces of laminated printed circuit boards, device/epoxy cement and heat spreader (i.e., finned heat sink or heat pipe), and any other metallic or non-metallic interstitial material employed between contacting interfaces.

Published experimental data of potential coatings, adhesives, and elastomeric gaskets is presented that can improve the thermal contact conductance of contacting surfaces within microelectronic systems. In addition, recommendations for future analytical and experimental studies of the mnechanistic principles, which control thermal performance of interstitial materials, are discussed for non-uniform pressure distribution.

The paper presents crack propagation experiments on real flip chip specimens applied to reversible shear loading. Two specially designed micro testers will be introduced. The first tester provides very precise measurements of the force displacement hysteresis. The achieved resolutions have been I mN for force and 20 nm for displacement. The second micro tester works similar to the first one, but is designed for in-situ experiments inside the SEM. Since it needs to be very small in size it reaches only resolutions of 10 mN and 100nm, which is sufficient to achieve equivalence to the first tester. A cyclic triangular strain wave is used as load profile for the crack propagation experiment. The experiment was done with both machines applying equivalent specimens and load. The force displacement curve was recorded using the first micro mechanical tester. From those hysteresis, the force amplitude has been determined for every cycle. All force amplitudes are plotted versus the number of cycles in order to quantify the crack length. With the second tester, images were taken at every 10th … 100th cycle in order to locate the crack propagation. Finally both results have been linked together for a combined quatitive and spatial description of the crack propagation in flip chip solder joints.

The paper presents new material data of real flip chip solder joints. A testing apparatus was designed to perform reversible shear tests on flip chip joints. In contrast to previous test setups for similar purposes this tester provides an infinite stiffness, a very high precision force (1 mN resolution) and displacement (20 nm resolution) measurement. The experimental program included cyclic shear tests for elastic plastic material data as well as a newly developed reversal creep and relaxation test for time depended material properties. A subsequent FEM simulation was applied to evaluate experimental raw data and to determine the parameters of material models provided by ANSYSTM. A user defined creep model had been added to the source code, in order to receive a convenient solder model for FE-Analysis of SnPb37 flip chip solder joints under thermomechanical stresses in electronic packages. The results of this study shows that the material behavior of flip chip solder joints is much more like that of bulk samples with a comparable micro structure than it is commonly believed up to now, because of the previously published data.

Epoxy-based, particulate composites are commonly used in packaging applications for the microelectronics industry. For example molding compounds, die attach adhesives, and encapsulant resins are all polymer matrix, particulate-reinforced composites. The mechanical behavior of these materials is a function of the type of resin, the shape and size of the reinforcing particles, the filler content, etc. In this paper, we will examine the interactions of the toughening mechanisms between rubber particles and the reinforcing particles. Hybrid-particulate composites are prepared by compounding both rubber particles and disc-shaped boron nitride particles in a bisphenol A-based epoxy cured with piperidine. Interestingly, there are combinations of rubber and reinforcing particles that lead to synergistic toughening. The underlying causes for synergistic toughening will be discussed.

An important phenomenon in silver (Ag) particle-filled adhesives is the development of electrical conductivity as the polymer composite is cured. We report the results of optical and electrical measurements performed on a commercially available Ag particle/polymer composite as a function of sample temperature. Surface enhanced Raman scattering (SERS) was used to probe the chemical nature of the Ag interface while a four-point contact probe monitored the onset of DC electrical conductivity. Complementary to the die attach adhesives studies, SERS experiments were also performed on commercially available neat Ag flake with controlled adsorbate coverages. For both the neat flake and filled adhesive, a carboxylate layer was attached to an oxygen-covered Ag flake surface via the carboxylic acid end of the molecule. The SERS results observed a partial decomposition of the carboxylate species into an amorphous carbon layer upon an increase in the temperature of both the flake and filled adhesive samples. However, the temperature threshold for the chemical conversion was lower for the Ag-filled adhesive relative to the neat flake. In the composite samples, the formation of the amorphous carbon layer occurred well below the curing temperature and coincided with a corresponding decrease in the electrical resistivity of the adhesive. Thus, an initial step in developing electrical conductivity appears to be the partial conversion of the lubricant adsorbate to an amorphous carbon layer.

Using an improved version of the thermal pulse (TP) method for determining electric field profiles and thermal parameters, we have obtained for the first time field profiles in 1.7 μm thick polyimide (PI) films spin-coated on Si wafers. As-coated films were found to have an internal field caused by negative charges within 200 nm of the free (pulsed) surface believed due to electrochemical reactions involving water. The change near the pulsed surface was stable up to 200 °C and would not be switched by applied fields of ±1 V/μm. The charge distribution and the thermal parameters were found to be sensitive to humidity