The diffusion barrier effect of Fe-Ni UBM as compared to the commercial Cu UBM during high temperature storage

Gao, Li-Yin; Li, Caifu; Wan, Peng; Zhang, Hao; Liu, Zhi-Quan

doi:10.1016/j.jallcom.2017.12.328

Cited by 21 publications

(4 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…17 These Fe-Ni alloys with 55 wt% to 70 wt% Fe are termed "Invar alloys". 17 The Invar Fe-Ni alloy has attracted researcher interest, 15,16,18,19 because of the possibility of reducing the thermal stress by matching the alloy's CTE with that of silicon and other substrate materials for integrated circuit packaging. However, the thermalexpansion properties and structural changes of the electroless-plated Fe-Ni-P alloy films during annealing, and which will be associated with the thermal-stress generation, have not been studied.…”

mentioning

confidence: 99%

Thermal Expansion and Thermal Stress Behavior of Electroless-Plated Fe–Ni–B Alloy Thin Film for High-Density Packaging

Yamamoto

Nagayama

Nakamura

2018

J. Electrochem. Soc.

View full text Add to dashboard Cite

Fe-Ni-B alloy thin films were prepared by using an electroless plating method, and their thermal-expansion behavior, thermal-stress generation, and crystal structure were evaluated. By heating from 30 • C to 300 • C, the electroless-plated Ni-5 wt%B alloy film showed a marked increase in tensile stress at approximately 270 • C, which is caused by a phase separation from the amorphous Ni-B alloy into a highly crystalline Ni and Ni 3 B compound. The coefficient of thermal expansion (CTE) of electroless-plated Ni-B alloy film from 300 • C to 30 • C, after heating to 300 • C, was 14 ppm/K. In contrast, the electroless-plated Fe-Ni-B alloy film in the Invar composition range had a small CTE (8 ppm/K to 9 ppm/K) and no substantial change in crystal structure during heating to 300 • C. As a result, the Fe-Ni-B alloy film on the silicon substrate (CTE = 3 ppm/K) exhibited a low thermal stress compared with the electroless-plated Ni-B alloy film. Thus, it is suggested that the electroless-plated Invar Fe-Ni-B alloy films are suitable as an under-bump metal layer from the viewpoint of the thermal-expansion behavior.

show abstract

mentioning

confidence: 99%

Thermal Expansion and Thermal Stress Behavior of Electroless-Plated Fe–Ni–B Alloy Thin Film for High-Density Packaging

Yamamoto

Nagayama

Nakamura

2018

J. Electrochem. Soc.

View full text Add to dashboard Cite

show abstract

“…The cross sections of different solder joints were prepared by the standard metallographic method as reported previously. 11,12 Furthermore, the grain size within the solder joints was analyzed by electron backscatter diffraction (EBSD). Prior to the EBSD observation, the cross-sectioned samples were first ground and polished mechanically and subsequently cleaned via an ion milling system (Leica EM RES101, Leica Microsystems, Germany) in order to obtain a high-resolution rate during observation.…”

Section: Characterization Methodsmentioning

confidence: 99%

Failure Mechanism of the SnAgCu/SnPb Mixed Soldering Process in a Ball Grid Array Structure

Gao

Cui

Tian

et al. 2020

Journal of Elec Materi

Self Cite

View full text Add to dashboard Cite

In this study, the influence of different reflow profiles and dosages of SnPb solder paste on the reliability of the SnAgCu/SnPb mixed assembly process was investigated. The interfacial microstructures were carefully examined through scanning electron microscopy, and two failures modes were found. Firstly, when reflowed at 245°C for 95 s, Pb-rich phases were buried within an intermetallic compound (IMC) layer and led to micro-voids and cracks under thermal stress or mechanical stress. However, upon increasing the reflow temperature or time further led to melting the solder ball totally, and the Pb element dissolved into a liquid phase and precipitated into uniform and small granules during the cooling process. Further, among the samples with different dosages of SnPb solder paste, severe shrinkage occurred at the solder joints with the addition of 12 wt.% eutectic SnPb component. Electron backscattered diffraction analysis was conducted in order to determine the mechanism of different shrinkages. The solder ball consisted of several primary Sn grains, and the shrinkage and Pb-rich phase are preferred to appear at the grain boundary. When the solder ball consisted of a single grain, no preferred orientation of the shrinkage and Pb-rich phase were detected. The shrinkage appeared at the final solidification region, which is near the IMC layer. Moreover, the shrinkages connect to each other, and a continuous crack formed on the chip side, leading to the failure.

show abstract

“…[1][2][3] As the packaging structure tends to be integrated and high density, the characteristic sizes of the interconnection, such as the dimensions of redistributed layer (RDL), pad, and under bump metallization, are gradually decreasing to the several micrometers. [4][5][6] With the small dimensions and the increase of stacking layers, there rise challenges on the mechanical properties and reliability of the electroplated copper materials. [7][8][9] The decrease of line width and pad diameter led to a stress concentration, which increased the risk of cracking failures.…”

Section: Introductionmentioning

confidence: 99%

Influence of Macroscale Dimension on the Electrocrystallization of Cu Pad and Redistributed Layer in Advanced Packaging

Liu,

Wan,

Gao

et al. 2024

Adv Eng Mater

Self Cite

View full text Add to dashboard Cite

Equiaxed copper (EG Cu), copper with columnar nanotwinned grains (C‐NT Cu), and copper with mixed equiaxed and columnar grains (MG Cu) were electroplated on pads and redistributed layers (RDLs) with different dimensions. When the diameters of pads increased from 40 μm to 105 μm, the flatness of pads decreased from 34.5% to 24.0%. And the flatness gradually deteriorated from 16.5 % to 34.4 % with the increase of RDL width from 8 μm to 20 μm. In contrast, the flatness of C‐NT Cu and MG Cu fluctuated between 1 % and 5 %, showing a weak correlation on the dimension. In terms of microstructure, the EG Cu and MG Cu had a weak correlation with size change, and the grain orientation showed a randomness. However, the (111) texture component of C‐NT Cu differed greatly with different dimensions, especially in transition layer. The (111) texture component changed from 9.6% to 38.1% for ϕ 40‐105 μm pad, and from 14.3% to 18.9% for 8‐20 μm wide RDL. The study reveals the significant size effect on the deposited profile and microstructure of electrodeposited copper materials, which should give insights of the materials design in advanced packaging technology.This article is protected by copyright. All rights reserved.

show abstract

The diffusion barrier effect of Fe-Ni UBM as compared to the commercial Cu UBM during high temperature storage

Cited by 21 publications

References 31 publications

Thermal Expansion and Thermal Stress Behavior of Electroless-Plated Fe–Ni–B Alloy Thin Film for High-Density Packaging

Thermal Expansion and Thermal Stress Behavior of Electroless-Plated Fe–Ni–B Alloy Thin Film for High-Density Packaging

Failure Mechanism of the SnAgCu/SnPb Mixed Soldering Process in a Ball Grid Array Structure

Influence of Macroscale Dimension on the Electrocrystallization of Cu Pad and Redistributed Layer in Advanced Packaging

Contact Info

Product

Resources

About