Warpage measurement of large area multitilted silicon substrates at various processing conditions

Citation
Sk. Bhattacharya et al., Warpage measurement of large area multitilted silicon substrates at various processing conditions, IEEE T COMP, 23(3), 2000, pp. 497-504
Citations number
11
Categorie Soggetti
Material Science & Engineering
Journal title
IEEE TRANSACTIONS ON COMPONENTS AND PACKAGING TECHNOLOGIES
ISSN journal
15213331 → ACNP
Volume
23
Issue
3
Year of publication
2000
Pages
497 - 504
Database
ISI
SICI code
1521-3331(200009)23:3<497:WMOLAM>2.0.ZU;2-F
Abstract
In order to find a low-cost solution for the future MCM-D packaging, a mult itiling approach through the incorporation of several tiles on a large carr ier substrate was studied. The multitiling format provides simultaneous pro cessing of several small silicon wafers on a carrier glass with a coefficie nt of thermal expansion (CTE) comparable to that of silicon. The wafers (ti les) are attached to the carrier glass (pallet) using a low modulus adhesiv e that can be released at an elevated temperature (similar to 450 degrees C ), The objective of this study is to develop materials and processes for a 12-in x 12-in (300-mm x 300-mm) large area substrate that can be scalable u p to a 600-mm x 600-mm format. The fabrication process begins with a carrie r CTE matched Borofloat glass on which silicon wafers are attached using a low modulus adhesive, This composite structure is exposed to high temperatu re thin-film processes that are required for the MCM-D manufacturing. The w arpage of these structures is a critical factor that determines the process ability of the thin films in a manufacturing environment, Specimen warpage was obtained using the shadow Moire technique. This noncontact, full field measurement method generates a moire pattern-a visual pattern of light and dark regions produced by the superposition of two regular motifs that geome trically interfere. A fringe pattern represents a "contour map" of the surf ace of the specimen. Warpage measurement was performed i) on as-received gl asses, ii) glasses after polishing, and iii) pallet assembly after tiling w as completed. Although polishing reduced the overall warpage of the as-rece ived pallets, the warpage of the tile and pallet assembly was increased aft er the adhesive was cured at 150 degrees C. Typical examples of warpage are 1.8 to 7.9 mils for the as-received glass, 2.7-3.0 mils for the polished g lass, and from 5.2-8.0 mils for the Borofloat assembly after curing. This p aper discusses the warpag issues associated with various stages of processi ng of the proposed large area MCM-D structures.