Effects of B or Al Interface Layers on Thermal Stability of Ni Silicide on Si

Abstract

A new method is proposed to address the degradation of thermal stability in Ni silicide on heavily doped N-type Si substrates. Layered deposition of Al or B at the Ni/Si interface prior to silicidation was found to improve the thermal stability of Ni silicide, as evaluated by the sheet resistance vs. silicidation temperature properties. The improvement is attributed to suppression of the agglomeration of silicide layers. An Al layer was effective only when it was located at the Ni/Si interface prior to the silicidation process. The use of a B layer was preferable to that of Al when considering lower junction leakage. IntroductionThe Ni silicidation process is an important process for state-of-the-art CMOS technology. However, problems with the thermal stability of Ni silicide have not yet been overcome completely. The thermal stability of Ni silicide is known to depend on impurities in the Si substrates (1). In particular, Ni silicide on heavily doped N + -Si substrates shows much inferior thermal stability than that on P + -Si substrates. The phenomenon indicates that acceptor impurities, i.e. group III elements, contribute to improvement of the thermal stability. Therefore, introduction of group III elements to the silicidation process without their electrical activation in Si is expected to improve the thermal stability of Ni silicide on N + -Si.A new method is proposed for reducing the degradation of thermal stability in Ni silicide on heavily doped N + -Si substrates, using the direct deposition of Al or B thin films at the Ni/Si interface. In this method, the N + region under the silicide layer is expected to be preserved due to the relative lowering of temperatures required during the silicidation process, which are not so high as to activate Al and B in Si. ExperimentalNi films were deposited in situ with additional Al or B films on chemically cleaned Si(100) substrates in various different layered structures, using an ultra-high vacuum (UHV) sputtering system. The total thickness of the stacked films for most samples was 12 nm, except for some samples with thick Al layers as described later. Heavily doped N + -Si substrates prepared by As implantation were mainly used, and lightly doped N -, lightly doped P -and heavily doped P + (B implanted)-Si substrates were also used as references. The deposited samples were annealed in forming gas (H 2 : 3%) at various temperatures for 60 s using a rapid thermal annealing (RTA) system. The formed silicide ECS Transactions, 13 (1) 413-419 (2008) 10.1149/1.2911524 © The Electrochemical Society 413 ) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 193.0.65.67Downloaded on 2015-06-08 to IP films were evaluated by sheet resistance measurements, Auger electron spectroscopy (AES) and scanning electron microscopy (SEM). Leakage currents of the N + /P junctions under the silicide layers were also measured. Results and discussionsThermal stability dependence on substrate dopingThe thermal stability ...