Reaction characteristics of the Au-Sn solder with under-bump metallurgy layers in optoelectronic packages

The reaction characteristics of molten Au/Sn eutectic solder with potential diffusion-barrier materials of optoelectronic packages were investigated. The characteristics were studied by reflowing the solder on Pt, Ni, and Co plates, respectively, and by measuring the thickness of the reaction product. In addition, the dissolution rate of Pt into the solder was measured. The results indicated that Pt, which is commonly used as the diffusion-barrier layer in the under-bump metallurgy (UBM) of optoelectronic packages, reacts readily with the molten solder, resulting in discontinuous reaction products at the solder/Pt interface. Cobalt, on the other hand, reacted with the solder at an order of magnitude slower rate than that of Pt and provided an effective barrier against the reaction with the solder.     

Fast vector-sum codebook search method for low bit rate speechcoding

A fast vector-sum codebook search method for low bit rate speech coding is presented. In this method, the codebook search is simplified by designing a vector-sum codebook that consists of orthonormal regular pulse basis vectors. A further simplification is achieved by adopting backward filtering. The method proposed has significantly reduced computational complexity, compared with the conventional VSELP, without producing any additional degradation in the quality of the synthesised speech     

A chemical mechanical polishing model incorporating both the chemical and mechanical effects

A comprehensive model for the material removal in a chemical mechanical polishing (CMP) process is presented in which both chemical and mechanical effects are taken into consideration. The chemical effects come into play through the formation of chemically modified surface layer on the wafer surface that, in turn, is removed mechanically by the plastic deformation induced by slurry particles. This model describes the influence of most, if not all, variables involved in the CMP process including slurry characteristics (solid loading, particle size and distribution, modulus), pad properties (modulus, hardness, asperity sizes and distribution) and processing conditions (down-pressure, velocity). Although more elaborate experimental verification of the model is yet to follow, this model appears to be capable of explaining many experimental observations on both oxide and metal CMP that, otherwise, could not be explained properly.     

Model-based analysis of the silica glass film etch mechanism in CF4/O2 inductively coupled plasma

The analysis of the Er-doped silica glass films (62%SiO₂–30%B₂O₃–8%P₂O₅ + 0.2 wt%. Er₂O₃) etch mechanism in the CF₄/O₂ inductively coupled plasma was carried out using the combination of simplified models for plasma chemistry and etch kinetics. As the O₂ mixing ratio in the CF₄/O₂ plasma increases from 0% to 30%, the etch rate decreases monotonically in the range of 385–190 nm/min that contradicts with the behavior of F atom density and flux. From the model-based analysis, it was found that, at low ion bombardment energies, the etch process followed the formal kinetics of ion-assisted chemical reaction and was controlled by both neutral and ion fluxes.     

Characterization of GaN and In x Ga1−x N films grown by MOCVD and MBE on free-standing GaN templates and quantum well structures

Structural and optical studies have been performed on GaN, InGaN layers, In_0.08Ga_0.92N/GaN heterostructures, In_0.08Ga_0.92N/In_0.02Ga_0.98N single and multiquantum wells grown by metal organic chemical vapor deposition (MOCVD) and GaN by molecular beam epitaxy (MBE) on GaN templates by using transmission electron microscopy (TEM), X-ray diffraction (XRD), and photoluminescence (PL). The layers are found to be high quality with low defect density, on the order of 10⁶ cm^?2, which are mainly related to the threading dislocations originating/propagating from the hydride vapor phase epitaxy (HVPE) GaN template. The interface between the layers and substrate could not be detected by TEM and was therefore deemed to be of high quality. Convergent beam electron diffraction studies revealed that the polarity of the films is Ga-polarity, which is the same as that of the substrate. A dual structure with different compositions and having thicknesses of 10 and 25 nm was observed in InGaN layers grown on GaN in one of the heterostructure samples. The full width at half maximum (FWHM) of the XRD rocking curves of (0 0 0 2) for heterostructures and quantum wells were found to be in the range of 15–28 arcmin for a slit width of 2 mm. PL studies on GaN layers grown by MBE and MOCVD on GaN templates are reasonably similar. The PL spectra from all the MBE and MOCVD epilayers and the substrate contain a plethora of sharp peaks related to excitonic transitions. With the presence of donor-bound exciton peaks and their associated two-electron satellites, the binding energies of two distinct shallow donors (28.8 and 32.6 meV), which are attributed to Si and O, respectively, were determined. PL measurements revealed that the FWHM of the main donor bound exciton peak increased from 0.6 to 2.9 meV but no change in peak position (3.472 eV) was observed in GaN when doping with Si (5×10¹⁷ cm^?3). However, the intensities of the yellow band and the shallow donor–acceptor pair band increased 10 times as compared to that in the undoped GaN samples. In the case of InGaN/GaN heterostructures, a similar trend was observed when compared to the doped samples. In the multiquantum well In_0.08Ga_0.92N/In_0.02Ga_0.98N heterostructures, the activation energy of the exciton emission, found to be 18 meV, was the lowest in the samples studied. The peak at 3.02 eV related to the InGaN was strongly pronounced in the In_0.08Ga_0.92N/In_0.02Ga_0.98N multiquantum well structure. In the In_0.08Ga_0.92N/In_0.02Ga_0.98N quantum well structures, the change in peak position with variation of temperature from 15 to 300 K in PL spectra is “S”-shaped. The cause for the “S” shape, i.e., a red–blue–red shift, is discussed.     

Development of artificial neural networks for real time, in situ ellipsometry data reduction

Development of real time in situ monitoring and control of thin film depositions using ellipsometry requires both data acquisition and processing to be rapid. Present speeds of measurement and computation of basic parameters, Ψ and Δ, are sufficient for data acquisition which is essentially real time. However, computation of film parameters, such as thickness and optical properties, generally cannot keep up with the incoming data and must be performed in a batch mode after the deposition.

This work describes the development of enhanced, high speed data reduction algorithms using artificial neural networks (ANN). The networks are trained using computed data and subsequently give values of film parameters in the millisecond time regime. The ANN outputs are used as initial estimates in a variably damped least squares algorithm for accuracy improvement. The combination of these two algorithms provides very accurate solutions in 75 ms per point on a DEC VAX 8800 multiprocessor system running at a combined 12 Mips. This speed is suitable for real time film monitoring and control for growth rates up to 10 nm per second. Results for fixed angle of incidence, single wavelength, in situ data for Ni deposited on BK7 substrates are presented.     

Healing behavior of a matrix crack on a carbon fiber/mendomer composite

The cure kinetics of a thermally mendable polymer based on Diels-Alder (DA) and retro-Diels-Alder (rDA) reactions, mendomer 401, was investigated using dynamic differential scanning calorimetry (DSC). The resulting behavior was modeled using a conventional cure kinetic model for thermosetting polymers. A composite panel with two layers of carbon fabric and mendomer 401 was fabricated following the cure cycle suggested by the cure kinetics model. Micro-indentation tests were performed to investigate mechanical properties and dynamic mechanical thermal analysis (DMTA) was conducted to study thermal behavior and to find the trigger temperature for healing. Self-healing behavior of the carbon fiber/mendomer composite was demonstrated using electrical resistive heating over the glass transition temperature.     

Self-sensing of carbon fiber/carbon nanofiber–epoxy composites with two different nanofiber aspect ratios investigated by electrical resistance and wettability measurements

Electro-micromechanical techniques, wettability test, and acoustic emission (AE) were use to compare self-sensing and stress-transferring effects in single carbon fiber embedded in carbon nanofiber (CNF)–epoxy composites with two different aspect ratios. Electrical resistivity and standard deviation were used as indirect measures of comparative dispersion degree of CNF. The dispersion was observed to decrease with increasing CNF content due to an increase in the electrical contacts. Composites with higher aspect ratio exhibited better self-sensing than lower aspect ratio case. This was attributed to differences in dispersion, orientation, coagulation of CNF with different aspect ratios. The opposite effect was observed for apparent Young’s modulus, which was larger for composites with lower aspect ratio. This is probably related to better stress transfer linked to orientation effects. Work of adhesion consistently followed same trend as apparent Young’s modulus. Single carbon fiber pull-out tests and AE provided additional information on the effects of aspect ratio.     

Observation of notch effect in Al6061-T6 specimens under tensile loading using digital image correlation and finite element method

The fracture of ductile materials such as Al6061-T6 usually starts from void nucleation. In this study, we investigated strengthening effect and stress triaxiality behavior in notched Al alloy specimens. Various specimens, one un-notched and two with different U-types of notch, one with large diameter (L-notched) and another with small diameter (S-notched), having the same width were selected. The distribution of strain in the full-field was examined using the digital image correlation (DIC) method. Under the tensile test, S-notched specimens showed a more pronounced strengthening effect than the L-notched specimens. The results of DIC confirmed that the notch tip of the S-notched specimen experienced high strain, while the L-notched specimen experienced high strain at the center. The maximum stress triaxiality moved rapidly from the notch tip to the center with the increase in notch radius over 10 mm. The results of the stress triaxiality analysis by finite element method were in good agreement with the above DIC measurement results and predicted the void generation and fracture sites in L-notched and S-notched specimens.