Integration of epitaxial LiNbO3thin films with silicon technology.

Development of bulk acoustic wave filters with ultra-wide pass bands and operating at high frequencies for 5thand 6thgeneration telecommunication applications and micro-scale actuators, energy harvesters and sensors requires lead-free piezoelectric thin films with high electromechanical coupling and compatible with Si technology. In this paper, the epitaxial growth of 36°Y-X and 30°X-Y LiNbO3films by direct liquid injection chemical vapour deposition on Si substrates by using epitaxial SrTiO3layers, grown by molecular beam epitaxy, has been demonstrated. The stability of the interfaces and chemical interactions between SrTiO3, LiNbO3and Si were studied experimentally and by thermodynamical calculations. The experimental conditions for pure 36°Y-X orientation growth have been optimized. The piezoelectricity of epitaxial 36°Y-X LiNbO3/SrTiO3/Si films was confirmed by means of piezoelectric force microscopy measurements and the ferroelectric domain inversion was attained at 85 kV.cm-1as expected for the nearly stoichiometric LiNbO3. According to the theoretical calculations, 36°Y-X LiNbO3films on Si could offer an electromechanical coupling of 24.4% for thickness extension excitation of bulk acoustic waves and a comparable figure of merit of actuators and vibrational energy harvesters to that of standard PbZr1-xTixO3films.

Direct liquid injection pulsed-pressure MOCVD of large area MoS2 on Si/SiO2.

Large-scale, high-quality growth of transition metal dichalcogenides (TMD) of controlled thickness is paramount for many applications in opto- and microelectronics. This paper describes the direct growth of well-controlled large area molybdenum disulfide (MoS2) on Si/SiO2 substrates by direct liquid injection pulsed-pressure metal-organic chemical vapor deposition (DLI-PP-MOCVD) using low-toxicity precursors. It is shown that control of the deposited thickness can be achieved by carefully tuning the amount of molybdenum precursor evaporated and that continuous layers are routinely obtained. Homogeneity and reproducibility have also been examined, as well as the average size of the grains. When targeting monolayer thickness, the MoS2 showed near stoichiometry (S/Mo = 1.93-1.95), low roughness and high photoluminescence (PL) quantum yield, equivalent to exfoliated monolayers and CVD MoS2 grown on the same substrates.

Heteroleptic Tin(IV) Aminoalkoxides and Aminofluoroalkoxides as MOCVD Precursors for Undoped and F-Doped SnO2 Thin Films.

A series of asymmetric and potentially bidentate amino alcohols and amino fluoro alcohols (RNOH) having a different number of methyl/trifluoromethyl substituents at the α-carbon atom, [HOC(R1)(R2)CH2NMe2] (R1 = R2 = H (dmaeH); R1 = H, R2 = CH3 (dmapH); R1 = R2 = CH3 (dmampH); R1 = H, R2 = CF3 (F-dmapH); R1 = R2 = CF3 (F-dmampH)) have been used to develop new monomeric and heteroleptic tin(IV) amino(fluoro)alkoxides [Sn(OR)2(ORN)2] (R = Et, Pri, But). These new complexes, which were thoroughly characterized by spectroscopy (IR and multinuclei NMR (1H, 13C, 19F, and 119Sn)) as well as single-crystal X-ray studies on representative samples, were investigated for their thermal behavior to determine their suitability as MOCVD precursors for the deposition of metal oxide thin films. The two most suitable compounds, [Sn(OBut)2(dmamp)2] and [Sn(OBut)2(F-dmamp)2], were used in a direct liquid injection chemical vapor deposition (DLI-CVD) process to deposit undoped SnO2 and F-doped SnO2 thin films, respectively, on silicon and quartz substrates. Film growth rates at different temperatures (from 400 to 700 °C), film thickness, crystalline quality, and surface morphology were investigated. The films deposited on quartz showed high transparency (above 80%) in the visible region and low carbon contamination on the surface (11-13% from XPS), which could easily be removed completely with 2 min of Ar+ sputtering.

Growth of low doped monolayer graphene on SiC(0001) via sublimation at low argon pressure.

Silicon carbide (SiC) sublimation is the most promising option to achieve transfer-free graphene at the wafer-scale. We investigated the initial growth stages from the buffer layer to monolayer graphene on SiC(0001) as a function of annealing temperature at low argon pressure (10 mbar). A buffer layer, fully covering the SiC substrate, forms when the substrate is annealed at 1600 °C. Graphene formation starts from the step edges of the SiC substrate at higher temperature (1700 °C). The spatial homogeneity of the monolayer graphene was observed at 1750 °C, as characterized by Raman spectroscopy and magneto-transport. Raman spectroscopy mapping indicated an AG-graphene/AG-HOPG ratio of around 3.3%, which is very close to the experimental value reported for a graphene monolayer. Transport measurements from room temperature down to 1.7 K indicated slightly p-doped samples (p ≃ 1010 cm-2) and confirmed both continuity and thickness of the monolayer graphene film. Successive growth processes have confirmed the reproducibility and homogeneity of these monolayer films.

Determining by Raman spectroscopy the average thickness and N-layer-specific surface coverages of MoS2 thin films with domains much smaller than the laser spot size.

Raman spectroscopy is a widely used technique to characterize nanomaterials because of its convenience, non-destructiveness, and sensitivity to materials change. The primary purpose of this work is to determine via Raman spectroscopy the average thickness of MoS2 thin films synthesized by direct liquid injection pulsed-pressure chemical vapor deposition (DLI-PP-CVD). Such samples are constituted of nanoflakes (with a lateral size of typically 50 nm, i.e., well below the laser spot size), with possibly a distribution of thicknesses and twist angles between stacked layers. As an essential preliminary, we first reassess the applicability of different Raman criteria to determine the thicknesses (or layer number, N) of MoS2 flakes from measurements performed on reference samples, namely well-characterized mechanically exfoliated or standard chemical vapor deposition MoS2 large flakes deposited on 90 ± 6 nm SiO2 on Si substrates. Then, we discuss the applicability of the same criteria for significantly different DLI-PP-CVD MoS2 samples with average thicknesses ranging from sub-monolayer up to three layers. Finally, an original procedure based on the measurement of the intensity of the layer breathing modes is proposed to evaluate the surface coverage for each N (i.e., the ratio between the surface covered by exactly N layers and the total surface) in DLI-PP-CVD MoS2 samples.