Browse Prior Art Database

Inverse Dual Damascene Pillar Structure for Chip-to-Package and 3D IC Connections

IP.com Disclosure Number: IPCOM000181687D
Original Publication Date: 2009-Apr-09
Included in the Prior Art Database: 2009-Apr-09
Document File: 4 page(s) / 35K

Publishing Venue

IBM

Abstract

The invention uses one or more, high aspect ratio Cu pillars with a wide base layer at the top and/or bottom of the pillar. The wide base layer eliminates the interface between the narrow pillar and the bond pad, thereby increasing the mechnical strength of the connection. The wide base at the bottom of the pillar is formed using a inverse dual damascene process. The wide base at the top of the pillar can be formed with or without an additional mask layer, by overfilling the pillar resist pattern. As a result, a triple damascene process is achieved with either one or two masks.Fig 1a. Dual damascene pillar.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 53% of the total text.

Page 1 of 4

Inverse Dual Damascene Pillar Structure for Chip -to-Package and 3D IC Connections

Copper pillar bumps are replacing solder bumps, because of improved electromigration reliability compared to solder. However, one problem with Cu pillars is that the stress in the on-chip wiring or at the solder-to-pillar interface can be higher compared to standard solder bumps, resulting in cracking during thermal cycle stress.

One solution to the cracking problem is to make high-aspect ratio Cu pillars, where the ratio of height : width is > 6:1 (E.B. Liao et al., EPTC 2005). However, the connections between the Cu pillars and the pads can be weak points where high stress occurs. Therefore, it is desirable to form Cu pillars with high area connections to the pads on the chip and the package.

The invention uses one or more, high aspect ratio Cu pillars with a wide base layer at the top and/or bottom of the pillar. The wide base layer eliminates the interface between the narrow pillar and the bond pad, thereby increasing the mechnical strength of the connection. The wide base at the bottom of the pillar is formed using a inverse dual damascene process. The wide base at the top of the pillar can be formed with or without an additional mask layer, by overfilling the pillar resist pattern. As a result, a triple damascene process is achieved with either one or two masks.Fig 1a. Dual damascene pillar.

Narrow standoff:
Sn extrusion minimized. Minimal Sn-Cu reaction Small solder ball size; small pitch Narrow standoff for flexibility

Wide base: EM resistant

BLM

Sn

Wide base:

Top view;

One or more narrow standoffs

Au polyimide

LM Cu

SiO2

Fig. 1b. Dual damascene pillar process. Form last metal...