Phase change random access memory appears to be the strongest candidate for next-generation high density nonvolatile memory. The fabrication of ultrahigh density phase change memory (≫1 Gb) depends heavily on the thin film growth technique for the phase changing chalcogenide material, most typically containing Ge, Sb and Te (Ge–Sb–Te). Atomic layer deposition (ALD) at low temperatures is the most preferred growth method for depositing such complex materials over surfaces possessing extreme topology. In this study, [(CH3)3Si]2Te and stable alkoxy-Ge (Ge(OCH3)4) and alkoxy-Sb (Sb(OC2H5)3) metal–organic precursors were used to deposit various layers with compositions lying on the GeTe2–Sb2Te3 tie lines at a substrate temperature as low as 70 °C using a thermal ALD process. The adsorption of Ge precursor was proven to be a physisorption type while other precursors showed a chemisorption behavior. However, the adsorption of Ge precursor was still self-regulated, and the facile ALD of the pseudobinary solid solutions with composition (GeTe2)(1‑x)(Sb2Te3) x were achieved. This chemistry-specific ALD process was quite robust against process variations, allowing highly conformal, smooth, and reproducible film growth over a contact hole structure with an extreme geometry. The detailed ALD behavior of binary compounds and incorporation behaviors of the binary compounds in pseudobinary solid solutions were studied in detail. This new composition material showed reliable phase change and accompanying resistance switching behavior, which were slightly better than the standard Ge2Sb2Te5 material in the nanoscale. The local chemical environment was similar to that of conventional Ge2Sb2Te5 materials.

Thermodynamic measurements on the Kondo lattice CeRhSn indicate a quantun critical point driven by geometrical frustration.

The magnetic field dependence of the spin-lattice relaxation rate 1/T 1 of 119 Sn in CeNiSn has been measured down to 20 mK in a field range of 0.2-8 T. (T 1 T) Ϫ1 is constant below 1 K, which depends on the external field in such a manner that its value stays constant to 2 T, while it increases linearly with the field up to 8 T. It is shown that this magnetic field dependence of (T 1 T) Ϫ1 is well explained by the simple scenario that the quasiparticle density of states at the Fermi level is produced by the Zeeman splitting of the up-and down-spin bands, keeping its V-shaped gapped structure unchanged for fields less than 8 T. The present experiment has elucidated that CeNiSn is in a semimetallic ground state with a low carrier density and the application of a magnetic field exceeding 2 T turns out to supply effective carriers. ͓S0163-1829͑96͒09933-X͔ PHYSICAL REVIEW B

No abstract

No abstract