Early Stage Of Deposition Research Articles

In Al N ∕ Ga N metal-oxide-semiconductor high electron mobility transistor with Al2O3 insulating films grown by metal organic chemical vapor deposition using Ar and NH3 carrier gases

Al 2 O 3 thin films were deposited by a metal organic chemical vapour deposition on InAlN∕GaN heterostructures using Ar or NH3 as a carrier gas. Effects of NH3 and Ar carrier gases on the electrical and structural properties of Al2O3∕InAlN∕GaN HEMT devices were investigated by current voltage, current collapse, and Auger electron spectroscopy measurements. Al2O3 deposited using Ar as a carrier gas leads to a substantial gate leakage current reduction with no increase of the current collapse compare to Schottky gate based HEMT. On the other hand, HEMT electrical performance shows degradation if NH3 carrier gas is used. Similarly, Auger electron spectroscopy revealed presence of carbon on InAlN surface when NH3 carrier gas was applied. It is suggested that the formation of carbon-related traps on the InAlN surface takes place in the early stages of the Al2O3 deposition with NH3 carrier gas that strongly influences the HEMT performance.

Electrocrystallisation of zinc from acidic sulphate baths; A nucleation and crystal growth process

The electrochemical nucleation and growth of zinc on low-carbon steel from acidic (pH 2.0–4.5) baths containing ZnSO 4, NaCl, and H 3BO 3, was studied by means of chronoamperometry at various cathodic potentials under a charge-transfer controlled regime. It is shown that at overpotentials in the range 0.30–0.55 V (negative to the Zn 2+/Zn redox value) the electrodeposition proceeds by instantaneous three-dimensional nucleation, which turns to progressive at higher overpotentials and/or very acidic baths. At low cathodic overpotentials (<0.30 V), a two-dimensional contribution limited by the incorporation of Zn ad-atoms in the developing lattice becomes significant at the early stages of deposition, and is more progressive in type the more acidic is the bath pH. Nucleation rate constants were calculated and correlated analytically with the respective potentials, using the classical theory of heterogeneous nucleation, which though fails to lead to reasonable values for the critical nucleus size.

Atomic and Electronic Structure at Au/CdSe Interfaces

By means of plane-wave pseudopotential periodic-supercell density functional theory calculations with a gradient-corrected exchange-correlation functional, we investigated the formation and the electronic structure of thin Au overlayers on CdSe(0001) and CdSe(0001) surfaces. We explored several possible Au/CdSe interfaces, including nonstoichiometric cases in which the very interface layer is mixed, namely, contains atoms of both the metal and the semiconductor. The relative formation energies of the computed model structures indicate that the formation of a very thin Au layer on CdSe surfaces can be epitaxial in the very early deposition stages but only in rather Au-rich conditions. The analysis of the band structures, densities of states, and wave functions for the low-energy interfaces reveals that hybridization occurs between the metal and the semiconductor electron states. This hybridization is confined at the very interface and is not expected to have significant consequences on the plasmonic and excitonic excitations that are appealing for nanotechnology applications of metal-semiconductor nanoparticles.

Molecular-dynamics investigation of the surface characteristics of Fe–Cu magnetic thin-film layers

Using molecular dynamics simulation, the structural characteristics of Fe and Cu thin films grown on Cu and Fe(001) substrates, respectively, were investigated with respect to the incident energy of adatoms and substrate temperature. In the case of Cu on Fe(001), no surface alloying at the interface was observed in the early stage of thin-film deposition, and growth generally followed the layer-by-layer growth mode. For Fe on a Cu(001) surface, a mixture confined to a single atomic layer at the Cu(001) surface was found to form at room temperature while films showed island growth. The steering effect due to atomic attraction was also observed at low incident energy, resulting in a rougher surface. Fe/Cu(001) growth changed to a layer-by-layer mode for an incident energy of 6 eV. The different aspects of surface morphology between Fe/Cu(001) and Cu/Fe(001) systems were explained in terms of surface free energy and impact cascade diffusion.

Study of the molecular beam epitaxial growth of InAs on Si-covered GaAs(1 0 0) substrates

The molecular beam epitaxial growth of InAs on GaAs (1 0 0) surfaces exposed to a Si-flux has been investigated. We varied the Si-coverage ( Θ Si) of the GaAs surface from 0 to 1.0 monolayers (ML), and studied the effects on the self-assembling of InAs nanoislands. It was found that for Θ Si equal to 0.3 and 0.6 ML, Si induces a bimodal size distribution that is associated to the creation of nucleation centers. We observed, that the inclusion of such nucleation centers changes the strain relaxation of the deposited InAs as compared with the sample grown with Θ Si=0. For Θ Si=1, we found that the three-dimensional (3D) islands start to nucleate from the very early stages of InAs deposition resembling the Volmer–Weber growth, contrasting with the usual Stranski–Krastanov growth mode of InAs on uncovered GaAs (1 0 0) substrates. For this sample the 3D growth mode is prompted by strain fields originated when Si is incorporated at the InAs–GaAs interface as an interlayer, as well as by the passivation of the GaAs surface by As–Si bonds. Atomic force microscopy images of the samples as well as reflectance difference spectroscopy also showed evidences of the InAs modified growth behavior. The GaAs interband transitions observed in the reflectance difference spectroscopy (RDS) spectra shifts in energy presumable due to changes in the strain status of the GaAs surface.

Nanostructure formation in InAs/InP(001) heteroepitaxy: Importance of surface reconstruction

Scanning tunneling microscopy has been used to study the deposition by molecular-beam epitaxy of thin InAs films on InP(001) substrates and compared with InAs growth on GaAs(001) under nominally identical conditions. In contrast to InAs/GaAs, InAs growth on InP does not proceed via a Stranski-Krastanov type mechanism, with a well-defined two-dimensional--three-dimensional growth mode transition, but instead a gradual and continuous surface roughening process occurs even from the earliest stages of deposition. Average height and surface roughness measurements indicate the absence of lateral surface correlations generated between adjacent elongated wires formed at higher deposition coverages. The origin of this growth behavior is attributed to the pregrowth formation of a self-templating arsenic-stabilized $\text{InP}(001)\text{\ensuremath{-}}(2\ifmmode\times\else\texttimes\fi{}4)$ surface, which is prepatterned on the atomic scale for the growth of highly anisotropic nanostructures.

Geochemistry and stable isotope composition of fresh alkaline porphyry copper tailings: Implications on sources and mobility of elements during transport and early stages of deposition

Prevention of acid mine drainage (AMD) in sulfide-containing tailings requires the identification of the geochemical processes and element pathways in the early stages of tailing deposition. However, analyses of recently deposited tailings in active tailings impoundments are scarce because mineralogical changes occur near the detection limits of many assays. This study shows that a detailed geochemical study which includes stable isotopes of water ( δ 2H, δ 18O), dissolved sulfates ( δ 34S, δ 18O) and hydrochemical parameter (pH, Eh, DOC, major and trace elements) from tailings samples taken at different depths in rainy and dry seasons allows the understanding of weathering (oxidation, dissolution, sorption, and desorption), water and element pathways, and mixing processes in active tailings impoundments. Fresh alkaline tailings (pH 9.2–10.2) from the Cu–Mo porphyry deposit in El Teniente, Chile had low carbonate (0.8–1.1 wt.% CaCO 3 equivalent) and sulfide concentrations (0.8–1.3 wt.%, mainly as pyrite). In the alkaline tailings water, Mo and Cu (up to 3.9 mg/L Mo and 0.016 mg/L Cu) were mobile as MoO 4 2− and Cu(OH) 2 0. During the flotation, tailings water reached equilibrium with gypsum (up to 738 mg/L Ca and 1765 mg/L SO 4). The δ 34S vs. δ 18O covariations of dissolved sulfate (2.3 to 4.5‰ δ 34S and 4.1 to 6.0 ‰ δ 18O) revealed the sulfate sources: the dissolution of primary sulfates (12.0 to 13.2‰ δ 34S, 7.4 to 10.9‰ δ 18O) and oxidation of primary sulfides (− 6.7 to 1.7‰ δ 34S). Sedimented tailings in the tailings impoundment can be divided into three layers with different water sources, element pathways, and geochemical processes. The deeper sediments (> 1 m depth) were infiltrated by catchment water, which partly replaced the original tailings water, especially during the winter season. This may have resulted in the change from alkaline to near-neutral pH and towards lower concentrations of most dissolved elements. The neutral pH and high DOC (up to 99.4 mg/L C) of the catchment water mobilized Cu (up to 0.25 mg/L) due to formation of organic Cu complexes; and Zn (up to 130 mg/L) due to dissolution of Zn oxides and desorption). At 1 m depth, tailings pore water obtained during the winter season was chemically and isotopically similar to fresh tailings water (pH 9.8–10.6, 26.7–35.5 mg/L Cl, 2.3–6.0 mg/L Mo). During the summer, a vadose zone evolved locally and temporarily up to 1.2 m depth, resulting in a higher concentration of dissolved solids in the pore water due to evaporation. During periodical new deposition of fresh tailings, the geochemistry of the surface layer was geochemically similar to fresh tailings. In periods without deposition, sulfide oxidation was suggested by decreasing pH (7.7–9.5), enrichment of MoO 4 2− and SO 4 2−, and changes in the isotopic composition of dissolved sulfates. Further enrichment for Na, K, Cl, SO 4, Mg, Cu, and Mo (up to 23.8 mg/L Mo) resulted from capillary transport towards the surface followed by evaporation and the precipitation of highly soluble efflorescent salts (e.g., mirabilite, syngenite) at the tailing surface during summer.

Characterization of Step-Edge Barriers in Organic Thin-Film Growth

Detailed understanding of growth mechanisms in organic thin-film deposition is crucial for tailoring growth morphologies, which in turn determine the physical properties of the resulting films. For growth of the rodlike molecule para-sexiphenyl, the evolution of terraced mounds is observed by atomic force microscopy. Using methods established in inorganic epitaxy, we demonstrate the existence of an additional barrier (0.67 electron volt) for step-edge crossing-the Ehrlich-Schwoebel barrier. This result was confirmed by transition state theory, which revealed a bending of the molecule at the step edge. A gradual reduction of this barrier in the first layers led to an almost layer-by-layer growth during early deposition stage. The reported phenomena are a direct consequence of the complexity of the molecular building blocks versus atomic systems.

Periodic Arrays of Cu-Phthalocyanine Chains on Au(110)

The structure of ultrathin Cu-phthalocyanine (Cu-Pc) films on the (1 × 2)-Au(110) surface has been studied. The overlayer deposition has been monitored in real time by helium atom scattering (HAS) and low energy electron diffraction (LEED). Throughout the monolayer regime the Cu-Pc molecules are systematically observed to line-up edge-to-edge along the [11̅0] direction of the Au substrate, yielding a commensurate 5-fold periodicity (14.4 Å). Cu-Pc chains deconstruct the 2-fold Au missing row order in the early stage of deposition. A set of higher order periodicities (5-, 7-, and 3-fold) are progressively observed along [001] with increasing Cu-Pc deposition, the 3-fold phase appearing at the monolayer saturation coverage. The corresponding molecular orientation has been studied by variable polarization absorption spectroscopy (XAS), whereas the Au substrate structure has been determined by out-of-plane surface X-ray diffraction. The (5 × 5) phase is found to be rather corrugated, and it exhibits a high degree of long-range order yielding the most prominent diffraction pattern. In the (5 × 5) phase, the Cu-Pc chains are found to lift the underneath missing row reconstruction, being separated by residual Au rows. Similarly, in the more compressed 3-fold monolayer phase, the Cu-Pc molecules were formerly found to lie within a shallow (1 × 3) Au reconstruction [Cossaro, A.; et al. J. Phys. Chem. B 2004, 108, 14671]. From comparison of the different deposition stages, as measured in real time by HAS, we can draw a comprehensive picture of the system evolution. In fact, the observed periodicities at different coverage are always formed by an array of Cu-Pc chains in shallow troughs that are equally spaced by a number of uncovered Au rows, as dictated by the Cu-Pc coverage. The growth of Cu-Pc arrays in the submonolayer range is thus driven by an interchain repulsion mechanism.

XPS studies of the ALD-growth of TaN diffusion barriers: Impact of the dielectric surface chemistry on the growth mechanism

The introduction of SiOCH low- k dielectrics in the copper interconnections of sub-45 nm node technologies is a challenge in terms of both material and process criteria. For instance, the deposition of a diffusion barrier between copper and dielectric is strongly dependent on the nature of the dielectric surface. In this study, we investigate the first steps of ALD TaN growth with respect to dielectric surface chemistry, using XPS measurements. Three different dielectrics have been tested: SiOCH, SiO 2 or SiOCH capped by a thin SiO 2 layer. We show that TaN can only grow over a monolayer-thick Ta 2O 5 formed at the early stages of deposition. A mechanism for the growth first steps is described, explaining the incubation delays observed for the appearing of Ta–O and Ta–N bonds. In addition, we show that a 3 nm-thick SiO 2 layer is able to hide the effect of SiOCH chemistry and obtain similar growth rates on SiOCH and SiO 2.

Development of an original model for the synthesis of silicon nanodots by Low Pressure Chemical Vapor Deposition

Using the Computational Fluid Dynamics code Fluent, a simulation model of an industrial Low Pressure Chemical Vapor Deposition reactor has been developed for the synthesis of silicon nanodots from silane SiH4 on silicon dioxide SiO2 substrates. A comparison between experimental and simulated deposition rates has shown that classical kinetic laws largely over-estimated these deposits. So, an original heterogeneous kinetic model is proposed as a first attempt to quantify the temporal evolution of deposition rates and of surface site numbers, as a function of operating conditions and of the chemical nature of substrate sites, for the early stages of silicon deposition. Contributions of silane and of the homogeneously born silylene SiH2 to nucleation and growth have been considered on different surface sites, silanol Si–OH, siloxane Si–O–Si and fresh silicon bonds. Simulations have revealed that for the conditions tested, the classical heterogeneous kinetic laws over-estimate, by more than 60%, silicon deposition during the first stages. The assumption that silylene and more largely all the unsaturated species formed in the gas phase contribute in priority to nucleation has been validated. Nucleation appears as a mandatory step to form the first fresh Si sites to allow deposition to occur from silane via growth phenomena.

Mechanism of growth and structure of titanium oxide ultrathin films deposited on Cu(0 0 1)

The growth mechanism, composition and structure of ultrathin films of titanium oxide deposited on the Cu(001) surface were investigated by means of XPS, LEIS, LEED and STM. Titanium oxide films were deposited on the Cu(001) surface previously saturated with a (√2×2√2)R45° structure of chemisorbed oxygen. The oxide films were prepared by evaporation of titanium in O2 atmosphere (pO2 in the 10−6mbar range) while the substrate temperature was kept at 573K. The Cu LEIS signal versus the amount of deposited titanium (as determined by means of XPS) indicates the growth of 2D islands in the early stages of deposition. Upon increasing the amount of deposited titanium multilayer islands begin to grow. The XPS results indicate that the oxide phases formed for Ti coverages above 1ML have a TiO2 stoichiometry. At very low coverages, a LEED pattern with a centred rectangular unit cell is observed. STM measurements show that at this stage of the growth the oxide islands are incorporated in the outermost layer of the substrate and the removed copper atoms form islands around the oxide regions. The very early stages of titanium oxide growth corresponding to the formation of this rectangular phase were also investigated by Ti deposition on the oxygen chemisorbed phase under UHV conditions at 573K. In this way it is possible to study the reaction of Ti with chemisorbed oxygen. Upon increasing the Ti coverage above 0.5ML, an oxide film with a slightly distorted hexagonal unit cell begins to grow. The quasi-hexagonal phase of titanium oxide can also be formed by annealing at 773K the rectangular phase. At higher coverages, when the substrate surface is completely covered by the oxide, the film exhibits a LEED pattern with a regular hexagonal unit cell.

Overpotential deposition of copper on an iodine-modified Au(1 1 1) electrode

Cyclic voltammetry, chronoamperometry and in situ electrochemical scanning tunneling microscopy were used to study the kinetics of nucleation and crystal growth during the initial stages of copper overpotential deposition (OPD) on a previously iodine-modified Au(1 1 1) electrode, from an aqueous solution 10 −3 M CuSO 4 in 0.05 M H 2SO 4. The starting potential during step experiments was chosen in the region where the gold electrode was completely free of the copper deposit. The recorded current transients for copper deposition onto the iodine-modified Au(1 1 1) electrode surface appear to be very complex, with the unusual presence of two or more current maxima. A new method was used for quantitative evaluation of current transients that involves the transition UPD–OPD, developed by our group [M. Palomar-Pardavé, I. González, N. Batina, J. Phys. Chem. B 104 (2000) 3545], was used for the quantitative interpretation. Our results show that, within a single current transient, copper adsorption and two types of nucleation process: two-dimensional (2D) and three-dimensional (3D) limited by lattice incorporation of copper adatoms and diffusion of Cu(II) ion, respectively, take place simultaneously. STM images revealed the enhanced growth of 3D copper on edge of I–Au(1 1 1) during the early stages of deposition. Moreover, our results strongly suggest that the iodine adlayer is constantly present, even after the striping Cu that was overpotential deposited.

Surface characterization of zincated aluminium and selected alloys at the early stage of the autocatalytic electroless nickel immersion process

In this work the changing structure of nickel–phosphorus deposits on aluminium and its alloys at the early stage of electroless nickel phosphorus deposition using hypophosphite ion as reducing agent has been studied. Prior to electroless nickel deposition, zincating is used for pre-treatment of aluminium substrates. The surface morphology and structure of the electroless Ni–P layers were characterized by scanning electron microscopy and X-ray diffraction analysis. Results show that Ni–P deposition is closely related to the dissolution of the zincating layer, followed by progressive nickel nucleation. The nuclei serve as a catalytic surface for further Ni–P deposition which increases with deposition time. The growth and coalescence of the nuclei on the aluminium substrate results in crystalline layers of Ni–P.

Structure of Nanowires Fabricated by Electron Beam Induced Deposition to Connect Self-Assembled Quantum Structures

Nanowires for measuring the electronic transport property of a self-assembled quantum dot and a quantum ring were fabricated by electron beam induced deposition (EBID), and cross-sectional transmission electron microscopy (TEM) observations were conducted for the first time for the nanowires. TEM observations revealed that an amorphous layer existed between the substrate and the W deposit, showing a low W deposition rate at the early stage of deposition, and confirmed that the layer was an amorphous carbon layer. The layer growth mechanism was discussed using a phenomenological theory considering two different precursors and assuming one of the precursors covers the entire surface initially. It was shown that the layer manner growth happens in a wide range of parameters. From high-resolution TEM, obvious signs of irradiation damage during the deposition were not observed, and the amorphous layer thickness was about 5 nm, which should not affect much transport property measurement, indicating that EBID is quite promising for nanowire fabrication for transport measurements of a single self-assembled nanostructure.

Oxidation of the Nb(1 1 0) surface by ab initio calculations

We report ab initio (FPLAPW) results referring to the early stages of oxygen deposition on the Nb(1 1 0) surface. We considered the cases of half and full monolayer depostion and we evaluated the structural and electronic properties of the resulting overlayers. From the calculated electronic strctures, it came out that the oxygen atoms form stronger bonds with the Nb surface atoms in the [0 0 1] than in the [ 1 1 ¯ 0 ] direction, a fact that implies also shorter bond lengths in the former direction and different structural relaxations in the neighbouring Nb atoms. In addition, we found that the Nb–O bonding, which is attributed to d–p hybridization of double bond-like character, results in the introduction of new electronic states that lay well below the lowest states of the clean Nb face, while it also contributes into the increase of states around the Fermi level. Interestingly, the Nb–O bonding type along the preponderated direction yields free oxygen p z orbitals that can be regarded as dangling-like bonds. These results elucidate the early stages of Nb oxidation and can be used to enlighten the mechanisms of surface reactions, catalytic activity and oxides growth mechanisms.

On the Origin of Mesoscopic Inhomogeneity of Conducting Polymers

The mesoscopic inhomogeneity of conducting polymer films obtained by electropolymerization and spin-coating was studied using Kelvin probe force microscopy (KFM) and current-sensing atomic-force microscopy (CS-AFM). A well-pronounced correlation was established between the polymer morphology, on the one hand, and its local work function (which is related to the polymer oxidation degree) as well as polymer conductivity, on the other. The most conducting regions were associated with the tops of the polymer grains and showed Ohmic behavior. They were surrounded first by semiconducting and then by insulating polymer. The conductivity of the grain periphery could be lower by as much as 2 orders of magnitude. The grain cores also showed consistently higher values of the local work function as compared to the grain periphery. This fact suggested that the grain cores were more oxidized and/or more ordered as compared to the grain periphery, which is in good agreement with the local conductivity data. More uniform morphology corresponded to less variability in the other properties of the polymer. A model is proposed that relates the observed inhomogeneity to preferential deposition of polymer molecules with higher molecular weight at the early stages of the polymer phase formation. The polymer deposition in either electropolymerization or various solution-casting techniques involves the nucleation of a new phase from a solution containing polymer fractions of different molecular weights. The driving force of the nucleation process depends on the solubility of the polymer fractions, which decreases with an increase in the molecular weight. This gives rise to preferential deposition of more crystalline, higher molecular weight polymer at the early stages of the polymer deposition to form the cores of the polymer grains. The fractions with lower molecular weights are deposited later and form less ordered/less conducting grain periphery. On the basis of this model, we conclude that, to ensure the formation of materials with low inhomogeneity and high quality, one should use the starting polymer with as narrow molecular weight distribution as possible. Yet another possibility is to use solvents which would reduce the differences in the solubilities of polymer fractions with different molecular weight.

The causes of apparent overconsolidation in the Namak marine deposit, Korea

An intensive experimental study was carried out to investigate the causes of the overconsolidation phenomena in the Namak New City site, Korea. Geological events in conjunction with sea level changes, groundwater fluctuation, weathering, leaching and cementation were examined as possible causes of overconsolidation. In addition to oedometer and triaxial tests, radiocarbon dating, geochemical components and pore water chemistry were analysed. It was shown that the overconsolidation in the upper marine clay deposit was not caused by stress reductions such as erosion and groundwater fluctuation. Furthermore, weathering or leaching seemed unlikely. The apparent overconsolidation can be explained by cementation, which is indicated by profiles of the sulphate and potassium ion concentrations in the pore water. At the early stage of the Holocene marine deposition, the rate of deposition was 4·4 mm/yr; thereafter it was reduced to 1·4 mm/yr, controlling chemical processes such as sulphate reduction and ion exchanges and, in turn, apparent overconsolidation. Some mechanical and chemical properties of the apparent overconsolidated clay were compared with the geologically overconsolidated Palaeosol layer in the site.

Electrodeposition of Al in 1-Butyl-1-methylpyrrolidinium Bis(trifluoromethylsulfonyl)amide and 1-Ethyl-3-methylimidazolium Bis(trifluoromethylsulfonyl)amide Ionic Liquids: In Situ STM and EQCM Studies

In the present paper, the electrodeposition of Al on flame-annealed Au(111) and polycrystalline Au substrates in two air- and water-stable ionic liquids namely, 1-butyl-1-methyl-pyrrolidinium bis(trifluoromethylsulfonyl)amide, [Py(1,4)]Tf(2)N, and 1-ethyl-3-methyl-imidazolium bis(trifluoromethylsulfonyl)amide, [EMIm]Tf(2)N, has been investigated by in situ scanning tunneling microscopy (STM), electrochemical quartz crystal microbalance (EQCM), and cyclic voltammetry. The cyclic voltammogram of aluminum deposition and stripping on Au(111) in the upper phase of the biphasic mixture of AlCl(3)/[EMIm]Tf(2)N at room temperature (25 degrees C) shows that the electrodeposition process is completely reversible as also evidenced by in situ STM and EQCM studies. Additionally, a cathodic peak at an electrode potential of about 0.55 V vs Al/Al(III) is correlated to the aluminum UPD process that was evidenced by in situ STM. A surface alloying of Al with Au at the early stage of deposition occurs. It has been found that the Au(111) surface is subject to a restructuring/reconstruction in the upper phase of the biphasic mixture of AlCl(3)/[Py(1,4)]Tf(2)N at room temperature (25 degrees C) and that the deposition is not fully reversible. Furthermore, the underpotential deposition of Al in [Py(1,4)]Tf(2)N is not as clear as in [EMIm]Tf(2)N. The frequency shift in the EQCM experiments in [Py(1,4)]Tf(2)N shows a surprising result as an increase in frequency and a decrease in damping with bulk aluminum deposition at potentials more negative than -1.8 V was observed at room temperature. However, at 100 degrees C there is a frequency decrease with ongoing Al deposition. At -2.0 V vs Al/Al(III), a bulk aluminum deposition sets in.

Early stage of Ge growth on Si(001) vicinal surfaces with an 8° miscut along[110]

The atomistic pathway towards the growth of semiconductor heterostructures on vicinal surfaces is investigated in a special experiment. A step-by-step study of the early stages of Ge deposition at T=600 degrees Celsius on a 8 degrees off Si(001) surface miscut along [110] is performed by scanning tunneling microscopy (STM). The microscopic processes occurring during growth are identified. Highly resolved STM images show how double height steps, which characterize the clean substrate, evolve by a step flow process generated by ad-dimer chains located at specific positions. This process leads to the formation of metastable single domains until the development of {105} faceted ripples extending along the whole surface in the miscut direction.