Assembly Strategies for 3D-IC Integration
With the advent of 3D Integration concepts as a potential enabler for the continuation of Moore’s law, the aligned wafer bonding requirements have shifted significantly. Chip-to-wafer bonding and wafer reconstruction represent alternative or complementary solutions to improve yield and chip size flexibility; pros and cons of the two methods are reviewed in the presentation made at Semicon Japan 2008.
3D assembly by chip-to-chip (C2C) and chip-to-wafer (C2W) stacking enables the development of very high density, multifunction devices and satisfies the demand for higher packaging miniaturization.
The trend toward higher density packaging means that chipmakers are being pressed to use the 3rd dimension to increase density for their chip packages.
Several 3D packaging techniques exist:
|
|
|
|
For chip stacking, the placement accuracy becomes more critical as the number of layers increases. The co-planarity of the layers is also an important factor. 3D considerations are crucial to assembly of MEMS devices, as well as other advanced 3D-structures, such as VCSEL arrays and smart pixel arrays. In all cases, 3D packaging requires unparalleled accuracy in the x-, y- and z-directions. S.E.T.'s Device Bonders provide this accuracy, with post-bond accuracy down to 0.5 µm and high leveling accuracy (20 µrad). All bonding techniques used for die attach and flip chip bonding can be applied: reflow, thermo-compression including Cu-Cu bonding, adhesive joining and even fusion bonding allowing for heterogeneous integration. Depending upon the interconnection density and the bonding technology used, picked and place machines or high accuracy die (flipped or not) bonders are used for attachment of the die to the wafer or to the previous die.
Download the pdf document.




