doi:10.1016/j.jcrysgro.2007.09.013
Copyright © 2007 Elsevier B.V. All rights reserved.
Effect of film thickness on the breakdown temperature of atomic layer deposited ultrathin HfO2 and Al2O3 diffusion barriers in copper metallization
Prodyut Majumdera, Rajesh Katamreddya and Christos Takoudisa, b,
, 
aDepartment of Chemical Engineering, University of Illinois at Chicago, Chicago, IL 60607, USA
bDepartment of Bioengineering, University of Illinois at Chicago, Chicago, IL 60607, USA
Received 10 August 2007;
revised 10 September 2007;
accepted 11 September 2007.
Communicated by D.W. Shaw.
Available online 19 September 2007.
References and further reading may be available for this article. To view references and further reading you must
purchase this article.
Abstract
We investigate the effect of thickness of HfO2 and Al2O3 barrier films on the breakdown temperature of Cu/barrier film/Si structures. The HfO2 and Al2O3 films are deposited using tetrakis-diethylamino hafnium and tris-diethylamino aluminum, respectively, as the metal precursors and ozone as the oxidizer at 250 °C. Interactions between the layers of Cu/barrier film (1 or 2 nm-thick)/Si structures due to high-temperature annealing in N2 are probed using sheet resistance measurements. The crystallinity of the multilayer structure and the possible formation of any anneal-induced reaction products are investigated with X-ray diffraction (XRD). The formation of Cu-silicide phase(s) due to diffusion of Cu atoms through the barrier layer indicates the failure of the corresponding diffusion barrier. The surface morphology of Cu is examined using scanning electron microscopy (SEM) and elemental mapping is done with energy dispersive X-ray spectroscopy (EDS). Our results show that both 1 and 2 nm-thick HfO2 and Al2O3 barrier films are capable of restricting the diffusion of Cu at high annealing temperatures; therefore, they could be used as effective diffusion barriers between Cu and Si.
Keywords: A1. Diffusion; A1. Nanostructures; A1. X-ray diffraction
PACS classification codes: 85.40.Ls; 66.30.Ny; 61.05.cp
Fig. 1. Variation of sheet resistance of Cu/HfO2 (1, 2 and 3 nm-thick)/Si as a function of annealing temperature. The samples are annealed for 5 min in 20 slpm of N2. (The sheet resistance of Cu/HfO2 (3 nm)/Si structures is cited from our paper [19].)
Fig. 2. Variation of sheet resistance of Cu/Al2O3 (1, 2 and 3 nm-thick)/Si as a function of annealing temperature. The samples are annealed for 5 min in 20 slpm of N2. (The sheet resistance of Cu/Al2O3 (3 nm)/Si structures is cited from our paper [19].)
Fig. 3. XRD patterns of Cu/barrier layer/Si samples with (a) 2 nm and (b) 1 nm HfO2, and (c) 2 nm and (d) 1 nm Al2O3. RTA is done for 5 min in N2 at the temperatures shown.
Fig. 4. SEM micrographs of the surface of Cu/HfO2 (2 nm-thick)/Si samples (a) as-deposited, and annealed at (b) 650 °C and (c) 675 °C. (d) EDS analysis of one of the crystallites shown in (c). SEM micrographs of the surface of Cu/HfO2 (1 nm-thick)/Si sample annealed at (e) 650 °C. RTA is done for 5 min in 20 slpm of N2.
Fig. 5. SEM micrographs of the surface of Cu/Al2O3 (2 nm-thick)/Si (a) as-deposited, and annealed at (b) 675 °C and (c) 700 °C. (d) EDS analysis of one of the crystallites in (c). SEM micrographs of the surface of Cu/Al2O3 (1 nm-thick)/Si sample annealed at (e) 675 °C. RTA is done for 5 min in 20 slpm of N2.