Multiple reflow study of ball grid array (BGA) solder joints on Au/Ni metallization

This paper evaluates the shearing behavior of ball grid array (BGA) solder joints on Au/Ni/Cu pads of FR4 substrates after multiple reflow soldering. A new Pb-free solder, Sn–3Ag–0.5Cu–8In (SACI), has been compared with Sn–3Ag–0.5Cu (SAC) and Sn–37Pb (SP) solders, in terms of fracture surfaces, shearing forces and microstructures. Three failure modes, ball cut, a combination of solder shear and solder/pad bond separation, and pad lift, are assessed for the different solders and reflow cycles. It is found that the shearing forces of the SP and SAC solder joints tend to increase slightly with an increase in the number of reflow cycles due to diffusion-induced solid solution strengthening of the bulk solder and augmentation of the shearing area. However, the shearing forces of the SACI solder joints decrease slightly after four cycles of reflow, which is ascribed to the thermal degradation of both the solder/intermetallic compound (IMC) and IMC/Ni interfaces. The SACI solder joints yield the highest strengths, whereas the SP solder joints give the smallest values, irrespective of the number of reflow cycles. Thickening of the interfacial IMC layer and coarsening of the dispersing IMC particles within the bulk solders were also observed. Nevertheless, the variation of shearing forces and IMC thickness with different numbers of reflow cycles was not so significant since the Ni under layer acted as an effective diffusion barrier. In addition, the initially-formed IMC layer retarded the further extensive dissolution of the pad material and its interaction with the solder

Modeling the diffusion of solid copper into liquid solder alloys

During the soldering process, the copper atoms diffuse into liquid solders. The diffusion process determines integrity and the reworking possibility of a solder joint. In order to capture the diffusion scenarios of solid copper into liquid Sn–Pb and Sn–Cu solders, a computer modeling has been performed for 10 s. An analytical model has also been proposed for calculating the diffusion coefficient of copper into liquid solders. It is found that the diffusion coefficient for Sn–Pb solder is 2.74 × 10− 10 m2/s and for Sn–Cu solder is 6.44 × 10−9 m2/s. The modeling results reveal that the diffusion coefficient is one of the major factors that govern the rate at which solid Cu dissolve in the molten solder. The predicted dissolved amounts of copper into solders have been validated with the help of scanning electron microscopic analysis.

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.

Vibrational spectroscopic studies of the structure of di-amino acid peptides. Part II: cyclo(L-Asp-L-Asp) in the solid state and in aqueous solution

Solid-state protonated and N,O-deuterated Fourier transform infrared (IR) and Raman scattering spectra together with the protonated and deuterated Raman spectra in aqueous solution of the cyclic di-amino acid peptide cyclo(L-Asp-L-Asp) are reported. Vibrational band assignments have been made on the basis of comparisons with previously cited literature values for diketopiperazine (DKP) derivatives and normal coordinate analyses for both the protonated and deuterated species based upon DFT calculations at the B3-LYP/cc-pVDZ level of the isolated molecule in the gas phase. The calculated minimum energy structure for cyclo(L-Asp-L-Asp), assuming C-2 symmetry, predicts a boat conformation for the DKP ring with both the two L-aspartyl side chains being folded slightly above the ring. The C=O stretching vibrations have been assigned for the side-chain carboxylic acid group (e.g. at 1693 and 1670 cm(-1) in the Raman spectrum) and the cis amide I bands (e.g. at 1660 cm(-1) in the Raman spectrum). The presence of two bands for the carboxylic acid C=O stretching modes in the solid-state Raman spectrum can be accounted for by factor group splitting of the two nonequivalent molecules in a crystallographic unit cell. The cis amide II band is observed at 1489 cm(-1) in the solid-state Raman spectrum, which is in agreement with results for cyclic di-amino acid peptide molecules examined previously in the solid state, where the DKP ring adopts a boat conformation. Additionally, it also appears that as the molecular mass of the substituent on the C-alpha atom is increased, the amide II band wavenumber decreases to below 1500 cm(-1); this may be a consequence of increased strain on the DKP ring. The cis amide II Raman band is characterized by its relatively small deuterium shift (29 cm(-1)), which indicates that this band has a smaller N-H bending contribution than the trans amide II vibrational band observed for linear peptides.

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.

Vibrational spectra of α-amino acids in the zwitterionic state in aqueous solution and the solid state: DFT calculations and the influence of hydrogen bonding

The zwitterionic forms of the two simplest alpha-amino acids, glycine and l-alanine, in aqueous solution and the solid state have been modeled by DFT calculations. Calculations of the structures in the solid state, using PW91 or PBE functionals, are in good agreement with the reported crystal structures, and the vibrational spectra computed at the optimized geometries provide a good fit to the observed IR and Raman spectra in the solid state. DFT calculations of the structures and vibrational spectra of the zwitterions in aqueous solution at the B3-LYP/cc-pVDZ level were found to require both explicit and implicit solvation models. Explicit solvation was modeled by inclusion of five hydrogen-bonded water molecules attached to each of the five possible hydrogen-bonding sites in the zwitterion and the integration equation formalism polarizable continuum model (IEF-PCM) was employed, providing a satisfactory fit to observed IR and Raman spectra. Band assignments are reported in terms of potential-energy distributions, which differ in some respects to those previously reported for glycine and l-alanine.

A vertex-based finite volume method applied to non-linear material problems in computational solid mechanics

A vertex-based finite volume (FV) method is presented for the computational solution of quasi-static solid mechanics problems involving material non-linearity and infinitesimal strains. The problems are analysed numerically with fully unstructured meshes that consist of a variety of two- and threedimensional element types. A detailed comparison between the vertex-based FV and the standard Galerkin FE methods is provided with regard to discretization, solution accuracy and computational efficiency. For some problem classes a direct equivalence of the two methods is demonstrated, both theoretically and numerically. However, for other problems some interesting advantages and disadvantages of the FV formulation over the Galerkin FE method are highlighted.

Dynamic model for metal cleanness evaluation by melting in a cold crucible

Melting of metallic samples in a cold crucible causes inclusions to concentrate on the surface owing to the action of the electromagnetic force in the skin layer. This process is dynamic, involving the melting stage, then quasi-stationary particle separation, and finally the solidification in the cold crucible. The proposed modeling technique is based on the pseudospectral solution method for coupled turbulent fluid flow, thermal and electromagnetic fields within the time varying fluid volume contained by the free surface, and partially the solid crucible wall. The model uses two methods for particle tracking: (1) a direct Lagrangian particle path computation and (2) a drifting concentration model. Lagrangian tracking is implemented for arbitrary unsteady flow. A specific numerical time integration scheme is implemented using implicit advancement that permits relatively large time-steps in the Lagrangian model. The drifting concentration model is based on a local equilibrium drift velocity assumption. Both methods are compared and demonstrated to give qualitatively similar results for stationary flow situations. The particular results presented are obtained for iron alloys. Small size particles of the order of 1 μm are shown to be less prone to separation by electromagnetic field action. In contrast, larger particles, 10 to 100 μm, are easily “trapped” by the electromagnetic field and stay on the sample surface at predetermined locations depending on their size and properties. The model allows optimization for melting power, geometry, and solidification rate.

The implementation of FT-Raman spectroscopy to study solid state moisture induced changes in the crystallographic nature of pharmaceutical excipients - a lesson in humidity

An aqueous solution of sucrose was lyophilised, producing amorphous sucrose. This wasthen stored under different humidity at 25ºC for 1 week, allowing some samples tocrystallise. FT-Raman spectroscopy and PXRD have been successfully shown toqualitatively distinguish between amorphous and crystalline samples of sucrose. The datafrom the two techniques is complementary.

The development of dynamic CFD model of the magnetic semi-levitation of liquid metals

In the casting of reactive metals, such as titanium alloys, contamination can be prevented if there is no contact between the hot liquid metal and solid crucible. This can be achieved by containing the liquid metal by means of high frequency AC magnetic field. A water cooled current-carrying coil, surrounding the metal can then provide the required Lorentz forces, and at the same time the current induced in the metal can provide the heating required to melt it. This ‘attractive’ processing solution has however many problems, the most serious being that of the control and containment of the liquid metal envelope, which requires a balance of the gravity and induced inertia forces on the one side, and the containing Lorentz and surface tension forces on the other. To model this process requires a fully coupled dyna ic solution of the flow fields, magnetic field and heat transfer/melding process to account for. A simplified solution has been published previously providing quasi-static solutions only, by taking the irrotational ‘magnetic pressure’ term of the Lorentz force into account. The authors remedy this deficiency by modelling the full problem using CFD techniques. The salient features of these techniques are included in this paper, as space allows.