Frequent failures in interconnected materials in microelectronic packages play a vital role in determining the reliability of electronic devices. Differences in the thermal expansion coefficients of the solder joint, printed circuit board (PCB), and electronic package are a major source of applied stress on solder joints. Due to health concerns and governmental regulations associated with using leaded solder alloys, SAC solder alloys are commonly used as an alternative to Pb alloys because of their outstanding solderability and reliability [organic solderability preservative (OSP)]. The shear fatigue creep accelerated test at different load levels (16, 20, and 24 MPa) was implemented in this study by using a universal microtesting machine to investigate the solder joint reliability. The creep effects were demonstrated by employing 60 s of dwell time for each alternating stress value and comparing its impact on solder joint reliability with the fatigue test results. The accelerated test was applied to SAC305 (96.5% Sn, 3% Ag, and 0.5% Cu) solder joints with OSP surface finish at different testing temperatures (25 °C, 60 °C, and 100 °C). Seven replicates were considered as a sample size for the reliability analysis. A two-parameter Weibull distribution was employed as the dominant distribution to describe the fatigue behavior at each condition. The characteristic life and shape parameters at each condition were extracted from the obtained Weibull models. The stress–life equation was used to estimate the reliability performance of the solder joints against fluctuations in stress amplitude at different testing temperatures. The inelastic work per cycle and plastic strain were calculated from the acquired hysteresis loop under each condition. Coffin–Manson and Morrow energy models illustrate the relationships between fatigue life versus plastic strain and inelastic work, respectively. The effects of the oscillated testing temperature on the stress–life model, Morrow energy model, and Coffin–Manson model were elaborated using the Arrhenius equation. The results indicated a severe degradation of fatigue life with increased testing temperature. Raising the stress amplitude had a less significant impact on fatigue life reduction than increasing the testing temperature. The reliability prediction models, as a function of the fatigue properties and testing temperature, were derived from the Arrhenius, Coffin–Manson, and Morrow energy models. The acquired prediction equations produced acceptable goodness-of-fit values. Finally, a general reliability model with a 99% adequacy value ( $R^{2}$ ) was constructed as a function of testing temperature and stress amplitude.

This article proposes a scalable substrate-integrated waveguide (SIW) module accommodating an in-line vector modulator monolithic millimeter integrated circuit (MMIC). The SIW module is realized with low-temperature co-fired ceramic (LTCC) technology, and it can be inserted in a dielectric-filled waveguide (DFWG). The module combines $\lambda _{g}/4$ -transformer-based $E$ plane tapering and SIWs on LTCC with the wire-bonded vector modulator. The proposed active LTCC module and two passive test structures (i.e., a constant-height-SIW module and a SIW module with $E$ plane taperings) are manufactured and tested as in-line modules in a DFWG. The passive test structures with the waveguide-to-DFWG and DFWG-to-SIW transitions measure 3.1 and 4.6 dB of insertion loss on average, respectively, at the 71–81 GHz frequency range. The active LTCC module measurements demonstrate a DFWG with phase and amplitude tuning capability and gain up to 17.6 dB within the same frequency range. A four-channel mock-up module with $\lambda _{0}/2$ channel spacing is designed and manufactured to demonstrate the scalability of the design.

This article demonstrates integrated packaging solutions for antenna components in D-band by using a glass-based package. We design sub-terahertz (sub-THz) patch antennas at 140 GHz frequency band using HFSS simulation and form arrays of patch antennas for 6G (sub-THz) wireless communication applications. The patch antennas and feeding networks are implemented using microstrip structures on the top of the interposer, with ground planes beneath. Build-up layers of dry polymer films are laminated on a glass core for multilayer copper metallization, and copper structures are patterned on the polymer layers. For precise fabrication of the antenna and feeding structures in a package, we implement the semi-additive patterning (SAP) process to deposit the copper structures. Adequate measurement methods are applied to address difficulties in $D$ -band antenna measurements. By forming patch antenna arrays, we obtain 10 dBi gain using a four-element linear array and 14 dBi gain using a 4-by-4 2-D array. The bandwidths achieved for these arrays are 7% (10 GHz) and 5% (7 GHz), respectively, based on return loss measurements. Overall, the measurement results present a good match to the simulation. Antenna structures on glass substrates demonstrated in this article represent the basic building block for heterogeneous integration of $D$ -band front-end module (FEM) in glass-based package.

A substrate-integrated waveguide (SIW) diplexer has been designed based on the dual-mode resonator operating with TE301 and TE103 modes in this letter. Techniques of harmonic interleaving, negative excitation and planar coupling structure, orthogonal transmission, and slot lines etched on the ground plane are exploited synthetically to extend the range of the out-of-band rejection. The diplexer with lower and upper channels centered at 34.95 and 36.44 GHz is fabricated and measured to verify the proposed concept, which exhibits a wide stopband better than 20 dB up to 82.92 GHz and a high isolation better than 28.9 dB over the whole band with three transmission zeros (TZs) nearby two passbands. Measurements accord well with the simulation results.

This article presents a comprehensive evaluation of 3-D-printed monolithic waveguide components fabricated by a high-precision micro laser sintering (MLS) process. The investigated devices are two 180-GHz bandpass filters and a straight $G$ -band (140–220 GHz) waveguide section. All were made of stainless steel, which was later gold coated using an electroless process. One of the filter samples was characterized using micro X-ray computed tomography (XCT) to inspect the printing quality as well as measure the internal dimensions. The sample was then sectioned to allow measurement of the surface roughness of the inner surfaces and inspect the gold coating quality. The as-manufactured stainless steel components showed high insertion losses: over 3 dB in the filter passbands and between 4.7 and 7 dB for the waveguide section, increasing with frequency over the $G$ -band. This loss is due to the electrical conductivity of stainless steel as well as the surface roughness. Gold plating significantly reduced the insertion losses to 0.5 dB for the filters and to between 0.6 and 1 dB for the waveguide section. The investigative study showed the high-dimensional accuracy and good printing quality achieved by MLS, demonstrating the value of the technique in producing monolithic metal waveguide components with fine geometrics.

Corrosion is a prevalent failure mode in electronic products. The initiation of failure often stems from preexisting corrosion contamination on soldering terminations prior to assembly. This corrosion is further accelerated by environmental factors such as humidity, temperature, and acidity, ultimately leading to the degradation of the board and failure during both postassembly testing and the product’s lifespan. This study presents a method for the real-time, early detection of corrosion contamination on electronic components during the mounting process using pick-and-place (PNP) technology. The method utilizes the correlation between light reflectance from soldering terminations during placement photography and the degree of corrosion present. Corroded terminations possess a rougher surface and pitting spots which result in different light reflectance compared to pristine terminations. This difference can be detected through artificial inteligence (AI) forensic analysis of component images. This study presents an AI model that correlates termination finish with corrosion content and progression, and evaluates its performance on large-scale data. This study also presents a real-world case where corroded components were identified during the PNP process, but later failed during in-circuit testing (ICT). The postfailure analysis, using scanning electron microscopy/energy-dispersive spectroscopy (SEM/EDS) and cross-sectional analysis, confirms the accuracy of the AI failure predictions on multiple components with corrosion, during large-scale production. The proposed method has been implemented in multiple production lines, where it inspects all components without compromising throughput, and identifies contaminated components that are unsafe. The method has been tested on over 3.5 billion components and has achieved an accuracy rate of over 99.5% in its predictions.

This article introduces a novel macromodeling method based on a recurrent neural network (RNN) called deep independently RNN (DIRNN). The proposed method applies to time-domain modeling of nonlinear circuits and components, resulting in better training. It overcomes the vanishing and exploding gradient problems encountered with conventional RNNs. In conventional RNNs, all neurons in each layer are involved in recurrent connections that cause unnecessary connections, increasing the model’s complexity over time and making it hard to train for long-time sequences. To solve this problem, the proposed DIRNNs neurons are independent of each other in recurrent connections because each neuron only receives connections from its own previous hidden state. The validity of the proposed method is verified by modeling two nonlinear circuit examples, namely, a multistage driver terminated by a multiline interconnect, and an ON-chip voltage generator.

This article analyzes the lithography design rules in package foundry and wafer foundry and reviews the major lithography techniques for package redistribution layer (RDL) fabrication for panel level 2.5-D/3-D interposers, fan-out packages, and heterogeneous integration. The techniques surveyed in this article are contact aligners, projection aligners/steppers, laser direct writing (LDW), and laser ablation, capable of resolving routing line and space (L/S) of 0.8– $1.5 \mu \text{m}$ with aspect ratio (AR) $\le 5$ and creating microvia with a diameter of 1.5– $2.0 \mu \text{m}$ at via pitch $\le 5 \mu \text{m}$ . The biggest challenge of advanced packaging is scaling the critical dimensions (CDs) on the package to keep up with the pace of scaling of the bump pitch. In addition, there is a critical need for patterning fine lines-and-spaces with high AR to have low resistance traces. These high AR ultra-fine lines and ultra-fine pitch vias are crucial for meeting the needs of high-bandwidth die-to-die interconnections at high input–output (I/O) densities. All these challenges are mainly driven by lithography. With the development of advanced photoresists (PRs), the resolution factor $K_{1}$ in projection lithography reduces from 0.66 to 0.39, improving the resolution by 40% without the negative impacts on the depth-of-focus (DOF). This article also discusses the specific lithography challenges associated with the topography of multi-layer RDL as well as their impacts on the fabrication of fine features. The fine pitch microvias can be a solution for scaling the I/O pitch down to 5– $10 \mu \text{m}$ as a bumpless way to connect copper pads of known-good-dies to known-good-substrates in fan-out packages.

In this study, the recent advances and trends in Cu–Cu hybrid bonding will be investigated. Emphasis is placed on the definition, kinds, advantages and disadvantages, challenges (opportunities), and examples of Cu–Cu bumpless hybrid bonding. Also, some recommendations will be provided.

This study proposes a method for the estimation of heat generation in semiconductor devices. The proposal is based on the inverse heat transfer problem (IHTP), where the objective function is minimized using Levenberg–Marquardt regularization, and the forward problem is solved with the 3-D finite-element method (FEM). The inverse problem is divided into two types of analyses. The first is a procedure named calibration, which is the IHTP for material properties and boundary condition estimation, and the second is the inverse problem for heat source estimation. The experimental subjects are a ball grid array (BGA), quad-flat-no-leads (QFN), and a heat-sink-small-outline package (HSOP). As a first step, the calibration procedure to estimate the heat capacity and thermal conductivity of a heater is carried. Said heater is used to provide a source term for the calibration of the experimental subjects. By placing the heater on the top surface of the mold, the unknown material properties and boundary conditions of each package are determined. After model calibration, the heat generation in each respective die is successfully estimated via inverse analysis, with a maximum error of 10% relative to the HSOP.