In recent years, novel electronic products like mobile phones, tablets, and personal computer have shrunk dramatically and become highly functionalized. To satisfy these market trends of smaller thinner devices, packaging technologies such as 3D-TSV, 2.5D, PoP and Flip-chip wafer bumping are being used. We have developed negative tone resists for re-distribution layer, C4 and micro bumps. Our resist is negative tone resist which incorporates an acrylate cross-linker system with photo radical initiator. This formulation shows excellent chemical resistance to various plating solutions such as Cu, Ni, Sn/Ag and Au, with good stripability. The key technology of plating resists for advanced packages is high resolution, allowing for high aspect ratio patterning capability that covers a wide range of film thicknesses. An example of such applications is in advanced PoP package design which require a thick film material that has high resolution capability, along with high aspect ratio performance. Additionally, high resolution re-distribution layer resists can be used to create fine pitch conditions which are necessary for package downsizing. Our resist shows good patterning performance from below 10 μm to above 100 μm film thickness conditions. In this paper, material design and key properties of novel plating resist with high resolution and high aspect ratio performance with a wide range film thickness ranges are discussed.

View full abstract

White LEDs with single phosphor usually have poor color rendering capability. The emission spectra of the blue LED and the yellow phosphor are narrow as compared with the reference light sources. The red light output is particular low. The color of the objects illuminated by such light source appear not accurate nor natural. Though the luminous efficiency is high, such poor color rendering light sources are not suitable for most general lighting applications. To improve the color rendering properties, multiple phosphors with different emission spectra should be used. For instance, a certain amount of red or orange phosphors may be used to mix with yellow phosphor to broaden the overall spectrum. In this paper, a numerical model is proposed to predict the emission spectra of LEDs with multiple phosphors. The model has considered the excitation and emission spectra of the phosphors, the mixing ratio and re-absorption between the phosphors. To validate the model, white light LEDs with multiple phosphors are fabricated. The spectra are measured and compared with the modeling results. It is found that the proposed model can estimate the emission spectra of LEDs with multiple phosphors with a high degree of accuracy.

View full abstract

The resistivity of Cu-TSVs (through silicon vias) substantially affects the performance of system LSIs. Hence, it is very important to evaluate the resistivity of Cu-TSV precisely.
Researchers and engineers have been concerned about the possibility that many impurities from additives and plating solutions were incorporated into Cu-TSV during plating and these impurities would precipitate along the grain boundaries, leading to great variation in grain sizes from tens of nm to several μm, resulting in a resistivity increase after annealing. However, the precise resistivity has not always been published. In order to obtain accurate resistivity, we have developed a new TEG (test element group) pattern and its manufacturing process. Using TEGs, we obtained 4.13 × 10^-8 Ω·m as the resistivity of one Cu-TSV. We discussed the mechanism for the high-resistivity of Cu-TSV based on measurements of grain size scattering using X-ray diffraction, EBSD (electron back scatter diffraction) and Cs-corrected STEM (scanning transmission electron microscope) observation.

View full abstract

In this study, the relationship between the residual stress and the warpage of Si chip flip-chip bonded (FCB) on a substrate using non-conductive film (NCF) is investigated with increasing temperature up to 120°C. The Si chip including piezo-resistive sensors for measuring the residual stress is diced to be 9 × 9 mm² size after thinned down to 200 μm or 550 μm thickness. Two different package substrates are used like as FR4 organic substrate and Si base one. The stress inside the chip is measured after FCB process using three kinds of NCFs, those have the different glass transition temperature (T_g). The residual stress decreased with increasing T_g of NCF for Si chip on organic substrate system regardless of Si chip thickness. Whereas, another package of Si chip bonded on Si substrate did not show a dependency on NCF. The warpages of Si chip bonded on each substrate are measured by Moiré method and the warpage maximum value and contours could be matched well to the measured residual stress values and shapes.

View full abstract

The present study discusses the formation of a conductive film from noble metal nanoparticles on a stainless steel substrate for use in the manufacture of electrical components, such as connectors. The proposed method consists of on-demand dispensing with nanoparticle paste followed by a brief preheating and laser sintering. The major results obtained are as follows: the laser sintering formed a gold film with a diameter of 0.8 mm and a thickness of 0.3-1.0 μm on the stainless steel substrate without any surface pretreatment; a laser with a wavelength of 915 nm enabled instantaneous sintering within one second in air; the laser-sintered gold nanoparticle film had such a high adhesion to the substrate that no separation occurred after 90°-0.5R bend-peel tests; the high adhesion was attributed to the interdiffusion of gold, iron, chromium and nickel in the course of sintering; primary sintering of the preheated gold nanoparticles at 523 K for 60 s with a small amount of solvents, and secondly sintering from the substrate side proceed simultaneously, making possible efficient sintering of the nanoparticles as well as high adhesion to the substrate; the laser-sintered gold film possessed a very good electrical property.

View full abstract

With non-destructive, simple, and low-priced testing methods (which can be applied to manufacturing processes), we can assure high quality production by applying inspection across all products. Quality assurance in wire bonding is possible by using state diagnosis which utilizes detection of elastic waves close to the bonded joint. We succeeded in detecting elastic waves under difficult conditions and in harsh environments by using a thin AE (Acoustic Emission) sensor developed using AlN piezoelectric thin film technology. We attempted to conduct a wire bonding state estimation using the MT (Mahalanobis-Taguchi) method which is a simple and powerful pattern recognition technique. Two issues are raised when applying this method to the manufacturing process: 1) a large probability bias caused by an insufficient number of data samples, and 2) securing homogeneity of the Unit Space. In this paper, we present a state diagnosis method utilizing a real-world application which uses piezoelectric thin film sensing and an improved MT method to which ensemble learning is applied. Performance of this proposed method is also examined through a common benchmark dataset to assure efficient operation in any manufacturing process.

View full abstract

High voltages and the edges of the metallization on ceramic substrates (AMB, DBA, DBC, HTCC, LTCC) lead to high electric field strengths. In the vicinity of the metal edges these high electric field strengths induce partial discharges in the ceramic insulation and in the potting and thereby represent one key degradation mechanism of power modules. In this work the correlation of the simulated electric field strength with phase resolved partial discharge (PRPD) measurements has been investigated. For the simulation of the electric field strength a new method was used to bypass numerical artifacts. The simulated values show that it is possible to reduce the electric field strength by an adaption of the metallization structure. There the distance of the upper and the lower metallization to the rim of the ceramic was changed relative to each other. Due to this variation a reduction of the electric field strength of about 30% can be reached by choosing the optimum distance compared to state of the art design. In PRPD measurements for insulating ceramic substrates (AlN/Al₂O₃ by DCB) we examined whether the electric field strength reduction leads to higher partial discharge inception voltages (PDIV). The measurements were executed on layouts with different dimensions of the upper and lower metallization relative to each other as well as for 3 different thicknesses of the ceramic insulation layer. An increase from 20% to 35% of the PDIV was measured for layouts designed according to the findings from the simulation with respect to electric field strength reduction. Finally, the calculated electric field strength and the measured PDIV were correlated.

View full abstract

Packaging material is one of the key components for through-silicon via (TSV) 2.5D/3D package giving a strong impact on the higher density integration and its reliability. In this paper, we report the assembly and reliability evaluation on film type underfill, photosensitive dielectric for redistribution layer as well as stress buffer, and molding compound. In the case of using our film type underfill, the chip is cracked by overflowed underfill during multi-die assembly in the conventional sequential method. By the collective staking method, the five-chip stacked package with TSV and 20 μm pitch bumps was demonstrated with excellent electrical yield. Also photosensitive dielectric realizes redistribution layers with 3 μm resolution and the raman shift in the 3D stack with photosensitive dielectric indicated lower stress than the one with silicon nitride as passivation layer. Finally we integrated our film type underfill, dielectric and molding compound in the stacked package, passing reliability tests such as thermal cycling and pressure cooker test.

View full abstract

Under the goal of prolonging lifetime of an electroless-copper-plating solution, a new method for supplying Cu²⁺ ions to the solution, copper-ball dissolution method was developed. The effect of the developed method was examined by comparing with conventional methods: CuO-particle dissolution method and CuSO₄-supply method. Studies on copper-ball dissolution reactions clarified that Cu²⁺ ion could be supplied to the plating solution by dissolving copper ball in the oxygen-supplied solution. Compared with CuSO₄-supply method, copper-ball dissolution method prevents salts from being accumulated in the solution, which can potentially double the lifetime of the solution, similar to CuO-particle dissolution method. The films obtained by copper-ball dissolution method contain fewer impurities than those deposited by CuO-particle dissolution. As a result, the mechanical properties of the deposited film and the reliability of through-hole connection are improved. It is concluded that copper-ball dissolution method is an effective method for prolonging the lifetime of an electroless copper-plating solution, leading to reduction in environmental loads.

View full abstract

A goal of quality assurance is to exhibit superior defect detection performance while minimizing the costs of inspection. In wire bonding, the acceptability of lots is decided based upon destructive test results of one or more samples drawn from the lot. To increase reliability of the inspection, it is necessary to conduct many destructive tests. Performance and costs are in a trade-off relationship, and thus it is desired to develop a method capable of improving the determining accuracy while preventing an increase in costs. For this reason, we focus on effective use of information obtained from "sample" as well as from "uninspected items". The former is accurate, but provides a relatively small dataset. Conversely, the latter is less accurate, but has an abundant dataset. The research described in this paper is an effort to improve classification performance of sampling inspection using the two concepts in a complementary way. In order to achieve highly accurate discrimination, we apply "Taming" - a machine learning problem - to the sampling inspection problem. Furthermore, sensing technology using a thin film Acoustic Emission (AE) sensor is implemented to evaluate the bonding process. The determining accuracy of the proposed method is compared with basic sampling inspection through actual bonding experiments.

View full abstract

In this paper, we investigate reliability testing for a glass interposer. The test vehicle includes a glass interposer with a chip substrate and a bismaleimide triazine (BT) substrate. The structure of a glass interposer with two redistribution layers (RDLs) on the front-side and one RDL on the back-side has been evaluated and developed. Key technologies, including via fabrication, front-side RDL formation, microbumping, temporary bonding, glass thinning, and back-side RDL formation, have been developed and integrated for high performance. The BT substrate design for electrical characterization of reliability tests is reported in this paper. The results indicate that this glass interposer can be effective for the assembly of thin substrates. The data shows the feasibility of this glass interposer for electronics applications.

View full abstract

This article was retracted.

View full abstract

This paper evaluates boiling heat transfer performance on a lotus-type porous copper plate attached onto a heated surface. The experiments are performed under atmospheric and saturated pool conditions. The lotus copper plates have uni-directional pore structure in a perpendicular direction for the heated surface and the average hydraulic diameter of the pore hole is 0.36 mm. The lotus copper plate of 1.0 mm or 2.0 mm in thickness is mechanically attached or soldered onto the heated surface in order to obtain the referential boiling heat transfer data of utilizing the lotus copper plate. The boiling curves suggest that the utilization of the lotus copper plate leads to boiling heat transfer enhancement especially in a low heat flux regime even when the lotus copper is mechanically attached to the heat transfer surface. However, the thickness of the lotus plate doesn’t affect the boiling heat transfer in the mechanically attached case, which indicates that there is a big contact thermal resistance between the heated surface and the lotus copper plate. In addition, it is confirmed that the critical heat flux in the mechanically attached case is almost the same as that in the smooth surface case. On the other hand, the critical heat flux increases up to 216 W/cm² by soldering the lotus copper plate to the heated surface, which is almost double of that in the smooth surface case.

View full abstract

This paper describes an improvement of a measurement method of thermal conductivity in order to shorten the measurement time. A novel thermal interface material (TIM) which has both high electrical insulation performance and high thermal conductivity is strongly demanded. Especially, the novel TIM is strongly needed to achieve the electrical insulation between inverters and water cooling jacket in EVs (Electric Vehicles) and HEVs (Hybrid Electric Vehicles). In order to shorten the development period of the TIMs that has to do many times of the experiment to optimize the composition of the TIMs, an improvement of an evaluation method of the thermal conduction performance of the TIMs should be prepared. However, steady state measurement methods of thermal conductivity generally take very long time. Therefore, we have been focusing on an improvement of measurement method of thermal conductivity in order to achieve the reduction of the measurement time. In this paper, we developed the fast thermal conductivity measurement method by predicting steady temperature from transient temperature history in equipment. Through the investigation, it is found that steady temperature rise in the steady state thermal conductivity measurement can be predicted by using the transient temperature history of the steady state measurement. Information about the optimum time period of the measurement time of the transient temperature response to predict the correct steady state temperature can be obtained.

View full abstract