The effects of the incident ion/metal flux ratio (1≤Ji /JTi≤15), with the N+2 ion energy Ei constant at ≂20 eV (≂10 eV per incident accelerated N), on the microstructure, texture, and stoichiometry of polycrystalline TiN films grown by ultrahigh-vacuum reactive-magnetron sputtering have been investigated. The layers were deposited in pure N2 discharges on thermally oxidized Si(001) substrates at 350 °C. All films were slightly overstoichiometric with a N/Ti ratio of 1.02±0.03 and a lattice constant of 0.4240±0.0005 nm, equal to that of unstrained bulk TiN. Films deposited with Ji/JTi=1 initially exhibit a mixed texture—predominately (111), (002), and (022)—with competitive columnar growth which slowly evolves into a pure (111) texture containing a network of both inter- and intracolumn porosity with an average column size of ≂50 nm at t=1.6 μm. In contrast, films grown with Ji/JTi≥5 do not exhibit competitive growth. While still columnar, the layers are dense with an essentially complete (002) preferred orientation and an average column size of ≂55 nm from the earliest observable stages. The normalized x-ray diffraction (002) intensity ratio in thick layers increased from ≂0 to 1 as Ji/JTi was varied from 1 to ≥5. Both 111 and 001 interplanar spacings remained constant as a function of film thickness for all Ji/JTi. Thus, contrary to previous models, strain is not the dominant factor in controlling the development of preferred orientation in these films. Moreover, once film texture is fully evolved—whether it be (002) or (111)—during deposition, changing Ji/JTi has little effect as preferred orientation becomes controlled by pseudomorphic forces. Film porosity, however, can be abruptly and reversibly switched by increasing or decreasing Ji/JTi.
Sputter-deposited Ti1−xWx diffusion barriers in microelectronic devices have been reported by many groups to be Ti deficient with respect to the target composition. In the present experiments, polycrystalline TixW1−x alloys were grown on oxidized Si(001) substrates at temperatures Ts between 100 and 600 °C by ultrahigh-vacuum magnetron cosputter deposition from pure W and Ti targets in 5 mTorr (0.65 Pa) Ar and Xe discharges. Films deposited in Ar were found by Rutherford backscattering and Auger electron spectroscopies to be increasingly Ti deficient with increases in the Ti sputtering rate and/or Ts at a constant W sputtering rate. TRIM calculations and Monte Carlo gas-transport simulations were used, in combination with the experimental results, to show that the Ti loss was due primarily to differential resputtering of the growing film by energetic Ar particles backscattered from the heavier W target. This effect is exacerbated at elevated film growth temperatures by Ti surface segregation in the alloy. The use of Xe, rather than Ar, as the sputtering gas greatly reduces both the flux and the average energy of backscattered particles incident at the substrate such that measurable Ti loss is no longer observed.
Polycrystalline bcc TixW1−x layers with mixed 011 and 002 texture were grown on oxidized Si(001) substrates at 600 °C by ultrahigh-vacuum (UHV) magnetron sputter deposition from W and Ti0.33W0.67 targets using both pure Ar and Xe discharges. Ti concentrations in the 100-nm-thick layers were 0, 6, and 33 at. % depending on target composition and sputtering gas. Al overlayers, 190 nm, thick with strong 111 preferred orientation, were then deposited in Ar at 100 °C with and without breaking vacuum. Changes in bilayer sheet resistance Rs were monitored as a function of time ta and temperature Ta during subsequent UHV annealing. Thermal ramping of Al/W and Al/Ti0.06W0.94 bilayers at 3 °C min−1 resulted in large (>fourfold) increases in Rs at Ta≃550 °C, whereas Rs in the Al/Ti0.33W0.67 bilayers did not exhibit a similar increase until ≃610 °C. Area-averaged and local interfacial reactions and microstructural changes were also followed as a function of annealing conditions. The combined results indicate that Al/W and Al/Ti0.06W0.94 bilayer reactions proceed along a very similar pathway in which monoclinic WAl4 forms first as a discontinuous interfacial phase followed by the nucleation of bcc WAl12 whose growth is limited by the rate of W diffusion, with an activation energy of 2.7 eV, into Al. In contrast, the W diffusion rate during the early stages of Al/Ti0.33W0.67 annealing is significantly higher allowing the formation of a continuous WAl4 interfacial blocking layer which increases the overall activation energy Ea, still limited by W diffusion, to 3.4 eV and strongly inhibits further reaction. We attribute observed increases in WAl4 nucleation and growth rates in interfacial Al/Ti0.33W0.67 to a “vacancy wind” effect associated with the very rapid (Ea=1.7 eV) diffusion of Ti into Al.
Interfacial reactions in epitaxial Al/TiN(111) model diffusion barriers: Formation of an impervious selflimited wurtzite-structure AIN(0001) blocking layer Single-crystal bee W(OO1) layers, 140 nm thick, were grown on MgO(001) substrates by ultrahigh-vacuum (UHV) magnetron sputter deposition at T,= 600 "C. Al overlayers, 190 nm thick with strong (001) and (011) preferred orientation and an average grain size of 200 nm, were then deposited at T,= 100 "C without breaking vacuum. Changes in bilayer sheet resistance R, were monitored continuously as a function of time t, and temperature T, during UHV annealing. In addition, Rutherford backscattering spectroscopy, x-ray diffraction, transmission electron microscopy (TEM), and scanning TEM, in which cross-sectional specimens were analyzed by energy-dispersive x-ray analysis with a 1 nm resolution, were used to follow area-averaged and local interfacial reaction paths as well as microstructural changes as a function of annealing conditions. The initial reaction products were discontinuous regions of monoclinic-structure WAl, which exhibit a crystallographic relationship with the underlying W layer. bee WA1i2 forms at a later stage and grows conformally to cover both W and WAl,. WAl, and WAl,, continue to grow, with W being the primary mobile species, until the Al layer is completely consumed. Information from the microchemical and microstructural analyses was used to model the R,(T, , ta) results based upon a multielement equivalent circuit approach which accounts for the observed nonplanar nature of the reaction front. Reaction kinetics and activation energies were determined. The results show that the growth of WAl, is diffusion limited with an activation energy E, of 3.1 eV while the formation of WAl,, is reaction limited with E,= 3.3 eV. 0 1995 American Institute of Physics.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.