Transfer of semiconductor and oxide films by wafer bonding and layer cutting |
| |
Authors: | Qin-Yi Tong Li-Juan Huang Ulich M. Gösele |
| |
Affiliation: | (1) Research Triangle Institute, RTP, 27709, NC;(2) School of Engineering, Duke University, 27708 Durham, NC;(3) Max-Planck Institute of Microstructure Physics, Weinberg 2, D-06120 Halle, Germany;(4) School of Engineering, Duke University, 27708 Durham, NC;(5) Present address: T.J. Watson Research Center, IBM, 10598 Yorktown Heights, NY |
| |
Abstract: | Material integration by wafer bonding and layer transfer is one of the main approaches to increase functionality of semiconductor
devices and to enhance integrated circuits (IC) performance. Even though most mismatches such as different lattice constants
betweeen bonding materials present no obstacle for wafer direct bonding, thermal stresses caused by thermal mismatches must
be minimized by low temperature bonding to avoid debonding, sliding or cracking. In order to achieve a strong bond at low
temperatures, two approaches may be adopted: 1) Bonding at room temperature by hydrogen bonding of OH, NH, or FH terminated
surfaces followed by polymerization to form covalent bonds. Within this approach the key is to remove the by-products of the
reaction at the bonding interface. 2) Direct formation of a covalent bond between clean surfaces without adsorbents in ultra
high vacuum conditions. Low temperature bonding allows bonding processed wafers for technology integration. Layer transfer
requires uniform thinning of one wafer of a bonded pair. The most promising technology involves a buried embrittled region
by hydrogen implantation. A layer with a thickness corresponding to the hydrogen implantation depth is then transferred onto
a bonded desired substrate by either splitting due to internal gas pressure or by forced peeling as long as the bonding energy
is higher than the fracture energy in the embrittled region at the layer transfer temperature. This approach is quite generic
in nature and may be applied to almost all materials. We have found that B+H co-implantation and/or H implantation at high
temperatures can significantly lower the splitting temperature. However, the wafer temperature during H implantation has to
be within a temperature window that is specific for each material. The experimentally determined temperature windows for some
semiconductors and single crystalline oxides will be given. |
| |
Keywords: | Wafer bonding film transfer materials integration |
本文献已被 SpringerLink 等数据库收录! |
|