Four examples of process capability improvemen t at WIN Semiconductors were demonstrated by using various statistical techniques. 0.35um pHEMT gate lithography process was improved by using Bayesian methodology to achieve more than 1.0μ μm DOF (depth of focus) and good CD control. A DOE with Taguchi analysis method was utilized to find the optimal implant isolation condition in HBT for achieving low and consistent isolation leakage current. By using the optimum metal thickness and alloy conditions, a low and stable HBT base Ohmic contact resistance was accomplished. DOE with ANOVA analysis have also been carried out for the scribe and break process to find the optimal scribing force and angle. All the optimized process has been implemented in WIN's production line to produce high performance and high yield HBT and pHEMT MMICs.

The effects of structural order on the light scattering characteristics of liquid crystal display cells containing smetic p-n-octyl-p′-cyanobiphenyl have been investigated. The scattering characteristics of different structures in the cell due to surface treatments are qualitatively analyzed for their effects on the contrast ratio when used in the thermally addressed liquid crystal display (TALC). Of the four structures investigated, i.e., homeotropic, fan-like, uniaxial rods and spherulites, the spherulitic structure has been found to exhibit the scattering profile best suited for high contrast because of the near-zero scattering near the incident beam. Based on the light scattering theory for spherulites, optimization of structural parameters such as size and order of spherulites, and device design such as collecting angle of projection lens for maximum display contrast has been predicted. Methods to improve contrast have also been discussed.

The design and operation of an automated light-scattering apparatus for studying structures in polymer films are described. The apparatus has angular resolution of 0.02 and 1.0° in radial and azimuthal angles, respectively. The radial scan can be made over the range of 0.5 (or less) to 90° and the azimuthal scan covers 180°. Time resolution is in the order of milliseconds. The operation, data collection, and reduction are fully automatic. Illustrative data on the crystallization of a low density polyethylene are described.

AbstractData on tensile strength and elongation at break for a series of Viton A‐HV vulcanizates are discussed. The data were obtained at various extension rates at temperatures from −5 to 230°C (25 ≲ T — Tg ≲ 260°C) on seven vulcanizates having crosslink densities ve (estimated from C1 in the Mooney‐Rivlin equation) from 0.46 × 10−5 to 24.4 × 10−5 mole/cm3. At an extension rate of 1 min−1, an increase in ve affects the tensile strength σb (based on the undeformed cross‐sectional area) and the true tensile strength σbσb (based on the cross‐sectional area of a deformed specimen) as follows: σb is essentially constant at a low temperature; it passes through a decided maximum at intermediate temperatures; and it increases to a plateau at elevated temperatures. In contrast, λbσb decreases markedly at all temperatures, an exception being the most lightly crosslinked vulcanizate(s). Application of time—temperature superposition to the ultimate‐property data gave log aT; its temperature dependence is that typical of nonpolar rubbery polymers. Data on the vulcanizates were compared in corresponding temperature states by plotting log 273σb/T, log 273λbσb/T, and (λb — 1)/(λb — 1)max against logtb/(tb)max, where tb is the temperature‐reduced time to break and (tb)max is the value at which the ultimate extension ratio λb attains its maximum, (λb)max. Except for the most lightly crosslink vulcanizate, the comparison shows that 273λbσb/T and (λb — 1)/(λb — 1)max are substantially independent of (or only weakly dependent on) crosslink density, that 273λb/T increases with ve, and that 273λb/T ∝︁ ve0.6 and λb ∝︁ ve−0.4 at a large value of tb/(tb)max.