2011
DOI: 10.1149/2.007202jes
|View full text |Cite
|
Sign up to set email alerts
|

Introduction of an Electroless-Plated Ni Diffusion Barrier in Cu/Sn/Cu Bonding Structures for 3D Integration

Abstract: This study examines the possibility of employing an electroless-plated Ni(P) layer as a diffusion barrier between the Sn bonding layer and Cu bump for 3D integration applications. We bonded the samples at different bonding temperatures (200∼350 • C) and probed into the bonding morphology to evaluate the effects of the addition of a Ni(P) barrier. Combination of scanning electron microscopy (SEM) and transmission electron microscopy (TEM) analyses revealed that the Ni(P) barrier effectively suppressed Cu diffus… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1

Citation Types

1
10
0

Year Published

2012
2012
2021
2021

Publication Types

Select...
7

Relationship

0
7

Authors

Journals

citations
Cited by 15 publications
(11 citation statements)
references
References 13 publications
(21 reference statements)
1
10
0
Order By: Relevance
“…The higher Cu concentrations and lower Sn concentrations observed below the electroless Ni-P plating layer as compared with those of the surroundings indicate the formation of Cu-Sn-based intermetallic compounds. Our FE-EPMA observations of the bonding interface of the p-type thermoelectric modules are consistent with previous studies, which concluded that Cu-Sn-based intermetallic compounds exist in the form of Cu 3 Sn and Cu 6 Sn 5 [23][24][25][26], and that they prevent interdiffusion between the Ni-P plating layer and Sn.…”
Section: Resultssupporting
confidence: 92%
See 2 more Smart Citations
“…The higher Cu concentrations and lower Sn concentrations observed below the electroless Ni-P plating layer as compared with those of the surroundings indicate the formation of Cu-Sn-based intermetallic compounds. Our FE-EPMA observations of the bonding interface of the p-type thermoelectric modules are consistent with previous studies, which concluded that Cu-Sn-based intermetallic compounds exist in the form of Cu 3 Sn and Cu 6 Sn 5 [23][24][25][26], and that they prevent interdiffusion between the Ni-P plating layer and Sn.…”
Section: Resultssupporting
confidence: 92%
“…The surface of the thermoelectric elements became rougher with increasing particle size of the alumina powder. As there was no chemical bonding at the interface between the thermoelectric element and the Ni-P plating layer, the adhesion of the electroless Ni-P plating layer could be secured via the physical surface roughness effect (i.e., the anchor effect) [23]. After sand blasting, the surfaces of thermoelectric element blocks subjected to electroless Ni plating were observed using a laser scanning confocal microscope ( Figure 2).…”
Section: Resultsmentioning
confidence: 99%
See 1 more Smart Citation
“…Because diffusivity data do not exist for Cu in the Ni(P) film, it is difficult to assess the validity of this proposition. However, the possibility is not considered high here, because the Ni(P) film is known to be a good diffusion barrier to Cu [22][23][24] and the current ENIG process temperature 358 K (85°C) was significantly lower than those used in other studies. [24] If the Cu diffusion to the Ni(P) film surface was primarily responsible for the black pad occurrence, the preannealing of the Cu UBM at 423 K (150°C) before the ENIG treatment should induce significantly more black pad formation, which contrast with the beneficial role of the pre-annealing reported in Reference 12.…”
Section: B Effects Of Ag Au Ni and Co Ubmsmentioning
confidence: 98%
“…There are reports on bonding schemes for 3D integration that deploy a thin Sn layer as an intermediate layer for Cu/Sn/Cu bump bonding structure. However, the temperature required for this technology is still within the range of 200-350 • C. 11,12 Here we report the method of using electrodeposited Ni cone layer in low-temperature bonding between Sn-capped Cu pillar bumps and substrate chip, as shown in Fig. 1a.…”
mentioning
confidence: 99%