A hybrid quantum mechanics (QM)/molecular dynamics (MD) simulation is performed to investigate the effect of an ionizable group (–SO₃⁻Na⁺) on polyaniline as gas sensing materials. Polymers considered for this work include emeraldine base of polyaniline (EB-PANI) and its derivatives (Na-SPANI (I), (II) and (III)) whose rings are partly monosubstituted by –SO₃⁻Na⁺. The hybrid simulation results show that the adsorption energy, Mulliken charge and band gap of analytes (CO₂ and H₂O) in polyaniline are relatively sensitive to the position and the amounts of –SO₃⁻Na⁺, and these parameters would affect the sensitivity of Na-SPANI/EB-PANI towards CO₂. The sensitivity of Na-SPANI (III)/EB-PANI towards CO₂ can be greatly improved by two orders of magnitude, which is in agreement with the experimental study. In addition, we also demonstrate that introducing –SO₃⁻Na⁺ groups at the rings can notably affect the gas transport properties of polyaniline. Comparative studies indicate that the effect of ionizable group on polyaniline as gas sensing materials for the polar gas molecule (H₂O) is more significant than that for the nonpolar gas molecule (CO₂). These findings contribute in the functionalization-induced variations of the material properties of polyaniline for CO₂ sensing and the design of new polyaniline with desired sensing properties.

Changes in the elasticity modulus of an epoxy molding compound (EMC), an electronic packaging polymer, under high-temperature air storage conditions, are discussed in this study. The elasticity modulus of EMC had two different compositions (different filling contents) under different temperatures (175, 200, and 225 °C) and aging times (100, 500, and 1500 h), which were analyzed by using dynamic mechanic analysis technology. The results revealed that the elasticity modulus increased in the thermal aging process, with an increase in the temperature and the aging time. The increments of the glassy and rubbery states were similar. However, the growing rate was significantly different, and the growth of the rubbery state was significantly higher than that of the glassy state. The filling content influenced the degree of aging of the materials significantly. At a low filling content, long-term aging under high temperatures completely changed the material structure, and the mechanical properties of the polymer were reduced.

Understanding the defect characterization of electronic and mechanical components is a crucial step in diagnosing component lifetime. Technologies for determining reliability, such as thermal modeling, cohesion modeling, statistical distribution, and entropy generation analysis, have been developed widely. Defect analysis based on the irreversibility entropy generation methodology is favorable for electronic and mechanical components because the second law of thermodynamics plays a unique role in the analysis of various damage assessment problems encountered in the engineering field. In recent years, numerical and theoretical studies involving entropy generation methodologies have been carried out to predict and diagnose the lifetime of electronic and mechanical components. This work aimed to review previous defect analysis studies that used entropy generation methodologies for electronic and mechanical components. The methodologies are classified into two categories, namely, damage analysis for electronic devices and defect diagnosis for mechanical components. Entropy generation formulations are also divided into two detailed derivations and are summarized and discussed by combining their applications. This work is expected to clarify the relationship among entropy generation methodologies, and benefit the research and development of reliable engineering components.

Bond wire damage is one of the most common failure modes of metal-oxide semiconductor field-effect transistor (MOSFET) power devices in wire-welded packaging. This paper proposes a novel bond wire damage detection approach based on two-port network measurement by identifying the MOSFET source parasitic inductance (L_S). Numerical calculation shows that the number of bond wire liftoffs will change the L_S, which can be used as an effective bond wire damage precursor. Considering a power MOSFET as a two-port network, L_S is accurately extracted from frequency domain impedance (Z−parameter) using a vector network analyzer under zero biasing conditions. Bond wire cutoff experiments are employed to validate the proposed approach for bond wire damage detection. The result shows that L_S increases with the rising severity of bond wire faults, and even the slight fault shows a high sensitivity, which can be effectively used to quantify the number of bond wire liftoffs of discrete MOSFETs. Meanwhile, the source parasitic resistance (R_S) extracted from the proposed two-port network measurement can be used for the bond wire damage detection of high switching frequency silicon carbide MOSFETs. This approach offers an effective quality screening technology for discrete MOSFETs without power on treatment.