Predicting optimal process conditions for flip-chip aassembly using copper column bumped dies

Recently, research has been carried out to test a novel bumping method which omits the under bump metallurgy (UBM) forming process by bonding copper columns directly onto the Al pads of the silicon dies. This bumping method could be adopted to simplify the flip chip assembly process, increase the productivity and achieve a higher I/O count. Computer modelling methods are used to predict the shape of solder joints and response of the flip chip to thermal cyclic loading. The accumulated plastic strain energy at the comer solder joints is used as the damage indicator. Models with a range of design parameters have been compared for their reliability. The ranking of the relative importance of these parameters is given. Results from these analyses are being used by our industrial and academic partners to identify optimal design conditions.

Computer modelling of the reliability of flip chips with metal column bumping

Recently, research has been carried out to test a novel bumping method which omits the under bump metallurgy forming process by bonding copper columns directly onto the Al pads of the silicon dies. This bumping method could be adopted to simplify the flip chip manufacturing process, increase the productivity and achieve a higher I/O count. This paper describes an investigation of the solder joint reliability of flip-chips based on this new bumping process. Computer modelling methods are used to predict the shape of solder joints and response of flip chips to thermal cyclic loading. The accumulated plastic strain energy at the comer solder joints is used as the damage indicator. Models with a range of design parameters have been compared for their reliability. The parameters that have been investigated are the copper column height, radius and solder volume. The ranking of the relative importance of these parameters is given. For most of the results presented in the paper, the solder material has been assumed to be the lead-free 96.5Sn3.5Ag alloy but some results for 60Sn40Pb solder joints have also been presented.

Lifetime prediction for power electronics module substrate mount-down solder interconnect

A numerical modeling method for the prediction of the lifetime of solder joints of relatively large solder area under cyclic thermal-mechanical loading conditions has been developed. The method is based on the Miner's linear damage accumulation rule and the properties of the accumulated plastic strain in front of the crack in large area solder joint. The nonlinear distribution of the damage indicator in the solder joints have been taken into account. The method has been used to calculate the lifetime of the solder interconnect in a power module under mixed cyclic loading conditions found in railway traction control applications. The results show that the solder thickness is a parameter that has a strong influence on the damage and therefore the lifetime of the solder joint while the substrate width and the thickness of the baseplate are much less important for the lifetime

Material properties, geometry and their effect on the fatigue life of two flip-chip models

This paper describes how modeling technology has been used in providing fatigue life time data of two flip-chip models. Full-scale three-dimensional modeling of flip-chips under cyclic thermal loading has been combined with solder joint stand-off height prediction to analyze the stress and strain conditions in the two models. The Coffin-Manson empirical relationship is employed to predict the fatigue life times of the solder interconnects. In order to help designers in selecting the underfill material and the printed circuit board, the Young's modulus and the coefficient of thermal expansion of the underfill, as well as the thickness of the printed circuit boards are treated as variable parameters. Fatigue life times are therefore calculated over a range of these material and geometry parameters. In this paper we will also describe how the use of micro-via technology may affect fatigue life

Reliability analysis of flip chip designs via computer simulation

A flip chip component is a silicon chip mounted to a substrate with the active area facing the substrate. This paper presents the results of an investigation into the relationship between a number of important material properties and geometric parameters on the thermal-mechanical fatigue reliability of a standard flip chip design and a flip chip design with the use of microvias. Computer modeling has been used to analyze the mechanical conditions of flip chips under cyclic thermal loading where the Coffin-Manson empirical relationship has been used to predict the life time of the solder interconnects. The material properties and geometry parameters that have been investigated are the Young's modulus, the coefficient of thermal expansion (CTE) of the underfill, the out-of-plane CTE (CTEz) of the substrate, the thickness of the substrate, and the standoff height. When these parameters vary, the predicted life-times are calculated and some of the features of the results are explained. By comparing the predicted lifetimes of the two designs and the strain conditions under thermal loading, the local CTE mismatch has been found to be one of most important factors in defining the reliability of flip chips with microvias.

Predicting the relationship between reliability and geometric parameters of cu column bumped flip-chips

Cu column bumping is a novel flip chip packaging technique that allows Cu columns to be bonded directly with the dies. It has eliminated the under-bump-metallurgy (UBM) fonnation step of the traditional flip chip manufacturing process. This bumping technique has the potential benefits of simplifying the flip chip manufacturing process, increasing productivity and the UO counts. In this paper, a study of reliability of Cu column bumped flip chips will be presented. Computer modelling methods have been used to predict the shape of solder joints and the response of flip chips to cyclic thermal-mechanical loading. The accumulated plastic strain energy at the corner solder joints has been used as an indicator of the solder joint reliability. Models with a wide range of design parameters have been compared for their reliability. The design parameters that have been investigated are the copper column height and radius, PCB pad radius, solder volume and Cu column wetting height. The relative importance ranking of these parameters has been obtained. The Lead-free solder material 96.5Sn3.5Ag has been used in this modelling work.

Computer modelling analysis of the globtop's effects on aluminium wirebond reliability

In power electronics modules, heavy aluminium wires, i.e. wire diameters greater than 100 microns, are bonded to the active semiconductor devices and conductor metallization to form electric circuits of the power electronic module. Due to the high currents that may flow through these wires, a single connection usually contains several wires and thus, a large number of wires are used in a power electronics module. Under normal operation or test condition, a significant amount of stresses and strains induced in the wire and bonding interfaces, resulting in failure over time. In this paper, computer modelling techniques are used to analyse the effect of globtop design on the reliability of aluminium wirebonds under cyclic thermal-mechanical loading conditions. The results will show the sensitivity of the reliability of the wirebonds to the changes in the geometry and the material properties of the wirebond globtop.

Computer simulation of crack propagation in power electronics module solder joints

A numerical modelling method for the analysis of solder joint damage and crack propagation has been described in this paper. The method is based on the disturbed state concept. Under cyclic thermal-mechanical loading conditions, the level of damage that occurs in solder joints is assumed to be a simple monotonic scalar function of the accumulated equivalent plastic strain. The increase of damage leads to crack initiation and propagation. By tracking the evolution of the damage level in solder joints, crack propagation path and rate can be simulated using Finite Element Analysis method. The discussions are focused on issues in the implementation of the method. The technique of speeding up the simulation and the mesh dependency issues are analysed. As an example of the application of this method, crack propagation in solder joints of power electronics modules under cyclic thermal-mechanical loading conditions has been analyzed and the predicted cracked area size after 3000 loading cycles is consistent with experimental results.

Computer simulation of aluminium wirebonds with globtop in power electronics modules

In this paper, computer modelling techniques are used to analyse the effects of globtops on the reliability of aluminium wirebonds in power electronics modules under cyclic thermal-mechanical loading conditions. The sensitivity of the wirehond reliability to the changes of the geometric and the material property parameters of wirebond globtop are evaluated and the optimal combination of the Young's modulus and the coefficient of thermal expansion have been predicted.

Vibrational spectroscopy and DFT calculations of di-amino acid cyclic peptides. Part I: cyclo(Gly-Gly), cyclo(L-Ala-L-Ala) and cyclo(L-Ala-Gly) in the solid state and in aqueous solution

Investigations of the vibrational spectra of cyclo(Gly-Gly), cyclo(L-Ala-L-Ala) and cyclo(t-Ala-Gly) are reported. Raman scattering and Fourier transform infrared (FTIR) spectra of solid-state and aqueous protonated samples, as well as their corresponding N-deuterated isotopomers, have been examined. In addition, density functional theory (DFT) (B3-LYP/cc-pVDZ) calculations of molecular structures and their associated vibrational modes were carried out. In each case, the calculated structures of lowest energy for the isolated gas-phase molecules have boat conformations. Assignments have been made for the observed Raman and FTIR vibrational bands of the cyclic di-amino acid peptides (CDAPs) examined. Raman polarization studies of aqueous phase samples are consistent with C-2 and C-1 symmetries for the six-membered rings of cyclo(L-Ala-L-Ala) and cydo(L-Ala-Gly), respectively. There is a good correlation between experimental and calculated vibrational bands for the three CDAPs. These data are in keeping with boat conformations for cydo(L-Ala-L-Ala) and cyclo(L-Ala-Gly) molecules, predicted by the ab initio calculations, in both the solid and aqueous solution states. However, Raman spectroscopic results might infer that cyclo(L-AlaGly) deviates only slightly from planarity in the solid state. The potential energy distributions of the amide I and II modes of a cis-peptide linkage are shown to be significantly different from those of the trans-peptides. For example, deuterium shifts have shown that the cis-amide I vibrations found in cyclo(Gly-Gly), cyclo(L-Ala-L-Ala), and cyclo(L-Ala-Gly) have larger N-H contributions compared to their trans-amide counterparts. Compared to trans-amide II vibrations, cis-amide II vibrations show a considerable decrease in N-H character.

IR/Raman spectroscopy and DFT calculations of cyclic di-amino acid peptides. Part III: comparison of solid state and solution structures of cyclo(L-Ser-L-Ser)

B3-LYP/cc-pVDZ calculations of the gas-phase structure and vibrational spectra of the isolated molecule cyclo(L-Ser-L-Ser), a cyclic di-amino acid peptide (CDAP), were carried out by assuming C-2 symmetry. It is predicted that the minimum-energy structure is a boat conformation for the diketopiperazine (DKP) ring with both L-Beryl side chains being folded slightly above the ring. An additional structure of higher energy (15.16 kJ mol(-1)) has been calculated for a DKP ring with a planar geometry, although in this case two fundamental vibrations have been calculated with imaginary wavenumbers. The reported X-ray crystallographic structure of cyclo(L-Ser-L-Ser), shows that the DKP ring displays a near-planar conformation, with both the two L-Beryl side chains being folded above the ring. It is hypothesized that the crystal packing forces constrain the DKP ring in a planar conformation and it is probable that the lower energy boat conformation may prevail in the aqueous environment. Raman scattering and Fourier-transform infrared (FT-IR) spectra of solid state and aqueous solution samples of cyclo(L-Ser-L-Ser) are reported and discussed. Vibrational band assignments have been made on the basis of comparisons with the calculated vibrational spectra and band wavenumber shifts upon deuteration of labile protons. The experimental Raman and IR results for solid-state samples show characteristic amide I vibrations which are split (Raman:1661 and 1687 cm(-1), IR:1666 and 1680 cm(-1)), possibly due to interactions between molecules in a crystallographic unit cell. The cis amide I band is differentiated by its deuterium shift of ~ 30 cm(-1), which is larger than that previously reported for trans amide I deuterium shifts. A cis amide II mode has been assigned to a Raman band located at 1520 cm(-1). The occurrence of this cis amide II mode at a wavenumber above 1500 cm(-1) concurs with results of previously examined CDAP molecules with low molecular weight substituents on the C-alpha atoms, and is also indicative of a relatively unstrained DKP ring.

Cyclic voltammetry and spectroelectrochemical study of nickel and cobalt diphenyltetraazaannulene complexes

The absorption spectra. cyclic voltammetry and spectroelectrochemistry of [Ni(II)DPTAA] and [Co(II)DPTAA] (DPTAA = 6,13-diphenyldibenzo[b,i][1,4,8,11] tetraaza[14]annulene) complexes in DMF are reported in detail. The ligand oxidation is observed for [Ni(II)DPTAA] at +0.70 V vs. SCE whereas Ni2(+/+) occurs at - 1.60 V. For [Co(II)DPTAA], a ligand oxidation redox couple is seen at +0.56 V while the Co2+/+ and Co2+/3+ redox couples appear at -1.21 and +0.24 V, respectively. All observed redox couples are assigned to reversible one-electron processes on account of peak separations and scan-rate dependency. These processes were further investigated by spectroelectrochemistry for [Co(II)DPTAA]. For [Co(II)DPTAA], axial ligation of pyridine was found to shift the Co2+/3+ redox couple more negative. while the ligand oxidation was shifted to more positive potentials. From a spectrophotometric titration of [Co(II)DPTAA] with pyridine an equilibrium constant, K-f, was determined for the binding of pyridine to [Co(II)DPTAA]. This was found to be 10.2 dm(3) mol(-1), slightly lower than that of [Co(II)TAA], indicating the influence of the phenyl groups. From this value and shifts in the Co2+/3+ redox couple upon ligation, an equilibrium constant for the binding of pyridine to [Co(III)DPTAA], K'(f), was found to be 5.06 x 10(6) dm(3) mol(-1). (c) 2007 Elsevier B.V. All rights reserved.