The reaction of amorphous Ni-Nb films with Si in the presence of a native SiO2 layer

Author(s):  
N. Rozhanski ◽  
A. Barg

Amorphous Ni-Nb alloys are of potential interest as diffusion barriers for high temperature metallization for VLSI. In the present work amorphous Ni-Nb films were sputter deposited on Si(100) and their interaction with a substrate was studied in the temperature range (200-700)°C. The crystallization of films was observed on the plan-view specimens heated in-situ in Philips-400ST microscope. Cross-sectional objects were prepared to study the structure of interfaces.The crystallization temperature of Ni5 0 Ni5 0 and Ni8 0 Nb2 0 films was found to be equal to 675°C and 525°C correspondingly. The crystallization of Ni5 0 Ni5 0 films is followed by the formation of Ni6Nb7 and Ni3Nb nucleus. Ni8 0Nb2 0 films crystallise with the formation of Ni and Ni3Nb crystals. No interaction of both films with Si substrate was observed on plan-view specimens up to 700°C, that is due to the barrier action of the native SiO2 layer.

2005 ◽  
Vol 20 (7) ◽  
pp. 1878-1887 ◽  
Author(s):  
Takanori Kiguchi ◽  
Naoki Wakiya ◽  
Kazuo Shinozaki ◽  
Nobuyasu Mizutani

The crystallization process of yttria-stabilized zirconia (YSZ) gate dielectrics deposited on p-Si (001) and SiOx/p-Si(001) substrates and the growth process of SiOx has been investigated directly using high-temperature in situ cross-sectional view transmission electron microscopy (TEM) method and high-temperature plan-view in-situ TEM method. The YSZ layer is crystallized by the nucleation and growth mechanism at temperatures greater than 573 K. Nucleation originates from the film surface. Nucleation occurs randomly in the YSZ layer. Subsequently, the crystallized YSZ area strains the Si surface. Finally, it grows in the in-plane direction with the strain, whereas, if a SiOx layer of 1.4 nm exists, it absorbs the crystallization strain. Thereby, an ultrathin SiOx layer can relax the strain generated in the Si substrate in thin film crystallization process.


Author(s):  
Hyoung H. Kang ◽  
Michael A. Gribelyuk ◽  
Oliver D. Patterson ◽  
Steven B. Herschbein ◽  
Corey Senowitz

Abstract Cross-sectional style transmission electron microscopy (TEM) sample preparation techniques by DualBeam (SEM/FIB) systems are widely used in both laboratory and manufacturing lines with either in-situ or ex-situ lift out methods. By contrast, however, the plan view TEM sample has only been prepared in the laboratory environment, and only after breaking the wafer. This paper introduces a novel methodology for in-line, plan view TEM sample preparation at the 300mm wafer level that does not require breaking the wafer. It also presents the benefit of the technique on electrically short defects. The methodology of thin lamella TEM sample preparation for plan view work in two different tool configurations is also presented. The detailed procedure of thin lamella sample preparation is also described. In-line, full wafer plan view (S)TEM provides a quick turn around solution for defect analysis in the manufacturing line.


1994 ◽  
Vol 9 (1) ◽  
pp. 63-67 ◽  
Author(s):  
G. Artioli ◽  
M. Bellotto ◽  
B. Palosz

A natural olivine sample from a mantle peridotite xenolith has been studied by in situ high-temperature powder diffraction. The structure has been successfully refined from powder data at three temperatures (25, 600, and 800 °C) using the Rietveld method. The study shows that the full-profile technique is well suited for the structure analysis of high-temperature powder diffraction data. The results indicate that, in this temperature range, there is no significant ordering of the Fe,Mg cations in the two crystallographically independent octahedral sites. This has implications for the thermodynamic modeling of olivine at upper mantle conditions. The present experiments allowed measurement of the lattice thermal expansion of olivine in the temperature range 25–800 °C, and assessment of the temperature dependence of the isotropic atomic displacement parameters.


1998 ◽  
Vol 248 (1-4) ◽  
pp. 109-114 ◽  
Author(s):  
Chih-Hao Lee ◽  
Hsin-Yi Lee ◽  
K.S. Liang ◽  
Tai-Bor Wu

1992 ◽  
Vol 280 ◽  
Author(s):  
I. Hashim ◽  
B. Park ◽  
H. A. Atwater

ABSTRACTEpitaxial Cu thin films have been grown on H-terminated Si(OOl) substrates at room temperature by D.C. ion-beam sputter deposition in ultrahigh vacuum. The development of orientation and microstructure during epitaxial growth from the initial stages of Cu growth up to Cu thicknesses of few hundred nm has been investigated. Analysis by in-situ reflection high energy electron diffraction, thin film x-ray diffraction, and plan-view and cross-sectional transmission electron microscopy indicates that the films are well textured with Cu(001)∥ Si(001) and Cu[100]∥ Si[110]. Interestingly, it is found that a distribution of orientations occurs at the early stages of Cu epitaxy on Si(001) surface, and that a (001) texture emerges gradually with increasing Cu thickness. The effect of silicide formation and deposition conditions on the crystalline quality of Cu epitaxy is also discussed.


Author(s):  
N. Rozhanski ◽  
V. Lifshitz

Pd-W alloy films are of interest because of their use for contact fabrication in microelectronic devices.In the present work Pd2 5W7 5, Pd4 0W6 0, Pd9 0W1 0 films were sputter deposited on Si(100) and only Pd9 0W1 0 films were polycrystalline.The amorphization of Pd- W films, possibly due to impurities, was not observed previously. The reaction of Pd9 0W1 0 films with a Si substrate begins at 200°C and leads to the formation of Pd2Si in two equivalent epitaxial orientations: A and B.In contrast to Ni-Nb films the presence of the native SiO2 layer under Pd-W alloy does not prevent the diffusion processes and the formation of silicides both on bulk and on plan-view specimens.Amorphous Pd-W films do not react with Si up to∼500°C and their crystallisation begins at T∼575°C.The reaction is also followed by the formation of two types of epitaxially oriented Pd2 Si islands as shown in Figure 1.The crystallisation of the films is accompanied by the formation of W grains.Pd crystals were not observed, that is possibly due to Pd precipitation at W grain boundaries.


1995 ◽  
Vol 402 ◽  
Author(s):  
André Vantommela ◽  
Stefan Degroote ◽  
Johan Dekoster ◽  
Hugo Bender ◽  
Guido Langouche

AbstractEpitaxial CoSi2(100) layers in the thickness range of 20 to 50 nm have been formed by reactive deposition epitaxy (i.e. Co deposition onto a hot Si substrate) without the use of either a template or an intermediate Ti layer. It is explained how growth parameters such as the deposition rate and substrate temperature are crucial in determining the epitaxial nature of the silicide. According to this model, good CoSi2/Si(100) alignment is only achieved when very low deposition rates are used (0.1 Å/s or less), combined with relatively high substrate temperatures during deposition (∼ 600°C or higher). Using these conditions, highly strained, continuous CoSi2 layers with a channeling minimum yield of χmin = 9% could be formed. Using higher rates and/or lower deposition temperatures, an increasing fraction of misoriented CoSi2 grains is presumed from backscattering/channeling and x-ray experiments, the nature of which is under investigation with plan view and cross sectional transmission electron microscopy.


1989 ◽  
Vol 148 ◽  
Author(s):  
Ki-Bum Kim ◽  
Robert Sinclair

ABSTRACTIn-situ annealing TEM experiments were performed on the Ti/GaAs system in order to study the dynamic behavior of interfacial reactions. Both plan-view and cross-sectional samples were investigated in either diffraction and imaging (both conventional and high resolution) modes. During experiments, we observed the following: (a) At the initial stage of reaction, the TiAs phase formed at the original Ti/GaAs interface with a distinct orientation with respect to the substrate; (b) as the reaction proceeded, the TiAs phase formed in a random manner; (c) finally, the liberated Ga species from the GaAs diffused out to the metal film and formed TiGa2 phase in the plan-view sample similar to the furnace-annealed case. For the cross-sectional sample, however, we did not observe any Ti:Ga phase formation. Instead, we observed the formation of voids both in the Ti film and in the GaAs substrate. The formation of different microstructure between in-situ and furnace annealed cases is explained by the sample geometry during annealing.


Sign in / Sign up

Export Citation Format

Share Document