AVS 57th International Symposium & Exhibition
    Thin Film Tuesday Sessions
       Session TF1-TuM

Paper TF1-TuM3
The Effects of Interfacial Organic Layers on the Growth of Thin Al2O3, HfO2 and TaNx Films by Atomic Layer Deposition

Tuesday, October 19, 2010, 8:40 am, Room Pecos

Session: ALD: Dielectrics for Semiconductors
Presenter: K.J. Hughes, Cornell University
Authors: K.J. Hughes, Cornell University
S. Issacson, Cornell University
J.R. Engstrom, Cornell University
Correspondent: Click to Email
One of the most poorly understood aspects of atomic layer deposition involves the initial stages of growth. As growth via ALD is almost always conducted on a substrate of different composition that the film being grown on top of it, there are often issues related to what are the best ways to initiate growth. Over the past few years we have been examining the use of interfacial organic layers (IOLs) as way to tailor the nucleation and growth of ALD thin films. Here, unlike essentially all work we have reported previously, we have conducted ALD using a conventional hot wall, viscous flow-type reactor, operating at pressures of ~ 1-100 mTorr. We focus here on the growth of two oxides: the benchmark Al2O3 and high-k HfO2, and one nitride, TaNx. As to the IOLs we consider one vapor phase deposited self-assembled monolayer, perfluorooctyltrichlorosilane (FOTS), which is expected to provide no active sites for ALD growth. We also consider a solution phase deposited polymer, poly(ethylene imine) (PEI ), which possesses a high density of –NH2 groups expected to act as active sites for growth. Concerning the latter IOL, from synchrotron x-ray reflectivity we confirm that PEI forms a thin (~ 7 Å) smooth film on the substrates we have examined. For ALD growth of Al2O3 on both bare SiO2, and SiO2 treated with PEI we observe linear growth, and no detectable incubation period. Growth on FOTS|SiO2 on the other hand clearly exhibits an incubation period of ~ 15 cycles. Thus, for Al2O3, growth on unmodified SiO2 and SiO2 modified with PEI are indistinguishable. In contrast, growth of both HfO2 and TaNx is definitely perturbed by the presence of the PEI interfacial organic layer: compared to growth on unmodified SiO2, growth on PEI |SiO2 exhibits an incubation period of ~ 10 cycles for HfO2, and > 40 cycles for TaNx. A common question is what is the fate of the IOL? Concerning the growth of Al2O3, from ex situ angle-resolved x-ray photoelectron spectroscopy (ARXPS) we find that the elemental species associated with the IOL (F for FOTS, and N for PEI ) are located at the IOL|Al2O3 interface, indicating the IOL has not migrated to the top surface during growth.