Properties of the SiO2/SiC interface investigated by angle resolved studies of the Si 2p and Si 1s levels and the Si KLL Auger transitions

Angle resolved photoemission studies of the Si 2p and Si 1s core levels and the Si KL_2,3L_2,3 Auger transitions from SiO₂/SiC samples are reported. Most samples investigated were grown in situ on initially clean and well ordered √3×√3 reconstructed 4H-SiC(0 0 0 1) surfaces but some samples were grown ex situ using a standard dry oxidation procedure. The results presented cover samples with total oxide thicknesses from about 5 to 118 Å. The angle resolved data show that two oxidation states only, Si⁺¹ and Si⁺⁴, are required to explain and model recorded Si 2p, Si 1s and Si KLL spectra.The intensity variations observed in the core level components versus electron emission angle are found to be well described by a layer attenuation model for all samples when assuming a sub-oxide (Si₂O) at the interface with a thickness ranging from 2.5 to 4 Å. We conclude that the sub-oxide is located at the interface and that the thickness of this layer does not increase much when the total oxide thickness is increased from about 5 to 118 Å.The SiO₂ chemical shift is found to be larger in the Si 1s level than in the Si 2p level and to depend on the thickness of the oxide layer. The SiO₂ shift is found to be fairly constant for oxides less than about 10 Å thick, to increase by 0.5 eV when increasing the oxide thickness to around 25 Å and then to be fairly constant for thicker oxides. An even more pronounced dependence is observed in the Si KLL transitions where a relative energy shift of 0.9 eV is determined.The relative final state relaxation energy ΔR(2p) is determined from the modified Auger parameter. This yields a value of ΔR(2p)=−1.7 eV and implies, for SiO₂/SiC, a “true” chemical shift in the Si 2p level of only ≈0.4 eV for oxide layers of up to 10 Å thick.     

Analysis of periodic Mo/Si multilayers: Influence of the Mo thickness

A set of Mo/Si periodic multilayers is studied by non-destructive analysis methods. The thickness of the Si layers is 5 nm while the thickness of the Mo layers changes from one multilayer to another, from 2 to 4 nm. This enables us to probe the effect of the transition between the amorphous and crystalline state of the Mo layers near the interfaces with Si on the optical performances of the multilayers. This transition results in the variation of the refractive index (density variation) of the Mo layers, as observed by X-ray reflectivity (XRR) at a wavelength of 0.154 nm. Combining X-ray emission spectroscopy (XES) and XRR, the parameters (composition, thickness and roughness) of the interfacial layers formed by the interaction between the Mo and Si layers are determined. However, these parameters do not evolve significantly as a function of the Mo thickness. It is observed by diffuse scattering at 1.33 nm that the lateral correlation length of the roughness strongly decreases when the Mo thickness goes from 2 to 3 nm. This is due to the development of Mo crystallites parallel to the multilayer surface.     

High-density-plasma (HDP)-CVD oxide to thermal oxide wafer bonding for strained silicon layer transfer applications

Direct wafer bonding between high-density-plasma chemical vapour deposited (HDP-CVD) oxide and thermal oxide (TO) has been investigated. HDP-CVD oxides, about 230 nm in thickness, were deposited on Si(0 0 1) control wafers and the wafers of interest that contain a thin strained silicon (sSi) layer on a so-called virtual substrate that is composed of relaxed SiGe (∼4 μm thick) on Si(0 0 1) wafers. The surfaces of the as-deposited HDP-CVD oxides on the Si control wafers were smooth with a root-mean-square (RMS) roughness of <1 nm, which is sufficiently smooth for direct wafer bonding. The surfaces of the sSi/SiGe/Si(0 0 1) substrates show an RMS roughness of >2 nm. After HDP-CVD oxide deposition on the sSi/SiGe/Si substrates, the RMS roughness of the oxide surfaces was also found to be the same, i.e., >2 nm. To use these wafers for direct bonding the RMS roughness had to be reduced below 1 nm, which was carried out using a chemo-mechanical polishing (CMP) step. After bonding the HDP-CVD oxides to thermally oxidized handle wafers, the bonded interfaces were mostly bubble- and void-free for the silicon control and the sSi/SiGe/Si(0 0 1) wafers. The bonded wafer pairs were then annealed at higher temperatures up to 800 °C and the bonded interfaces were still found to be almost bubble- and void-free. Thus, HDP-CVD oxide is quite suitable for direct wafer bonding and layer transfer of ultrathin sSi layers on oxidized Si wafers for the fabrication of novel sSOI substrates.     

Magnetic structures in Ni80Fe20/Ru multilayers

Using polarized neutron reflectivity (PNR) measurements together with associated simulation, magnetic structures of two Ni₈₀Fe₂₀ (1 1 1)/Ru (0 0 0 1) multilayer samples with Ru thickness of 9 and 21 Å were investigated under various external magnetic fields at room temperature. The results reveal the existence of layer thickness, interface roughness, magnetic moment, interlayer coupling angle and non-collinear coupling. The former three are independent of Ru thickness; while the latter two are strongly dependent of Ru thickness.     

XPS analysis with pulsed voltage stimuli

We record XPS spectra while applying 0 to +10 V or 0 to −10 V square pulses to the sample rod, which normally results in twinning of all peaks at correspondingly increased (for +10 V) or decreased (for −10 V) binding energies. For poorly conducting samples, like silicon oxide layer on a silicon substrate, the twinned peaks appear at different energies due to differential charging, which also vary with respect to the frequency of the applied pulses. Moreover, the frequency dependence varies with the thickness and can be correlated with the capacitance of the oxide layer. The technique is simple and can lead to extract important information related with dielectric properties of surface structures in a totally non-contact fashion.     

Optical losses and absorption cross-section of silicon nanocrystals

 Si-rich silicon oxide and SiO₂ (SRSO)/SiO₂ multilayer (ML) samples were grown by reactive magnetron sputtering and then annealed at high temperature to induce the formation of Si-nc with mean size of 3-4 nm and density of about 3.5×10¹⁸ cm⁻³ as deduced from high resolution TEM micrographs. Refractive index and thickness have been determined by m-line measurements, which have shown a birefringence of about 1.5% due to the ML structure. Rib-loaded waveguides have been fabricated to measure propagation losses in the visible-infrared range. The analysis of the different contributions to optical losses such as Mie scattering and scattering due to waveguide roughness has allowed us to isolate the contribution due to the absorption losses and thus to extract the absorption cross-section at different wavelengths. Values of about 3.5×10⁻¹⁸ cm² have been found at 830 nm, increasing with decreasing of the wavelength.     

Formation of Co ultrathin films on Si(1 1 1): Growth mechanisms, electronic structure and transport

The AES, EELS, AFM and resistance measurement investigations have been performed to determine the growth mechanism, electronic structure and resistance-thickness dependence of Co layers on silicon at the thickness range from submonolayer up to several monolayer coverage. These layers were obtained under UHV high-rate deposition with using re-evaporation of Co from a Ta foil. The layer-by-layer growth of Co on Si(1 1 1) with some light segregation of Si has been found on the AES data. An enlarged and reduced concentration of valence electrons in the interface Si layer at the thickness ranges 0-1 Å and in the Co film at d = 1-2 Å has been observed. Resistance measurement of the Co film showed a fast decrease of the resistance down to some value limited by quantum-size effect in accordance with the formation of a two-dimensional Co phase at d = 1-2 Å.     

Multiple antiferromagnet/ferromagnet interfaces as a probe of grain-size-dependent exchange bias in polycrystalline Co/Fe50Mn50

We have used ferromagnet/antiferromagnet/ferromagnet trilayers and ferromagnet/antiferromagnet multilayers to probe the grain size dependence of exchange bias in polycrystalline Co/Fe₅₀Mn₅₀. X-ray diffraction and transmission electron microscopy show that the Fe₅₀Mn₅₀ (FeMn) grain size increases with increasing FeMn thickness in the Co (30 Å)/FeMn system. Hence, in Co(30 Å)/FeMn(t_AF Å)/Co(30 Å) trilayers the two Co layers sample different FeMn grain sizes at the two antiferromagnet/ferromagnet interfaces. For FeMn thicknesses above 100 Å, where simple bilayers have a thickness-independent exchange bias, we are therefore able to deduce the influence of FeMn grain size on the exchange bias and coercivity (and their temperature dependence) simply by measuring trilayer and multilayer samples with varying FeMn thicknesses. This can be done while maintaining the (1 1 1) orientation, and with little variation in interface roughness. Increasing the average grain size from 90 to 135 Å results in a fourfold decrease in exchange bias, following an inverse grain size dependence. We interpret the results as being due to a decrease in uncompensated spin density with increasing antiferromagnet grain size, further evidence for the importance of defect-generated uncompensated spins.     

Reactivities of ultrathin alumina films exposed to intermediate pressures of H2O: Substrate-mediated mechanism for growth and loss of surface order

The response of ordered ultrathin Al₂O₃ films on NiAl(1 1 0) and Ni₃Al(1 1 0) substrates to sequential exposures at varying pressures of H₂O between 10⁻⁷ Torr and 10⁻³ Torr, ambient temperature, was characterized by LEED, AES and density functional theory (DFT) calculations. In all cases, an increase in average oxide thickness, as determined by AES, was observed, consistent with a field-induced oxide growth mechanism. Ordered oxide films of initial average thicknesses of 7 Å and 12 Å grown on NiAl(1 1 0) achieved a limiting thickness of 17(1) Å, while films of initial thickness of 7 Å and 11 Å grown on Ni₃Al(1 1 0) achieved a limiting thickness of 12(1) Å. The LEED patterns for the thinner (7 Å) films were not observed after exposure to 10⁻⁵ Torr (NiAl(1 1 0)), or 10⁻⁴ Torr (Ni₃Al(1 1 0)). In contrast, LEED patterns for the films of greater initial thickness persisted after exposures to 10⁻³ Torr UHV. DFT calculations indicate an Al vacancy formation energy that is significantly greater (by ∼0.5 eV) on the surface that has the thicker oxide film, directly opposite to what may be naively expected. A simple coordination argument supports these numerical results. Therefore, the greater limiting oxide thickness observed on NiAl(1 1 0) demonstrates that the rate determining step in the oxide growth process is not Al removal from the metal substrate and transport across the oxide/metal interface. Instead, the results indicate that the determining factor in the oxide growth mechanism is the kinetic barrier to Al diffusion from the substrate bulk to the oxide/metal interface. The persistence of the LEED patterns observed for the films of greater initial oxide thickness indicates that the surface disorder generally observed for alumina films grown on aluminide substrates and exposed to intermediate pressures of H₂O is due to the growth of a disordered alumina layer over an ordered substrate, rather than to direct H₂O interaction with terrace sites.     

Nanometric thinning of bonded silicon wafers using sacrificial anodic oxidation and investigated by X-ray reflectivity

X-ray reflectivity and atomic force microscopy are used to investigate silicon oxide ultra-thin films. Quantitative results are shown using a reflectivity simulation model based on kinematical X-ray theory. Changes in film thickness are discussed in relation to current density, voltage, charge and anodization time. The density and resistivity of silicon oxide are calculated and compared to that of thermal oxide. The electrical field existing in the layer during anodization is estimated. Surface roughness is also measured locally and averaged over the entire surface, producing a low value that meets microelectronic requirements. Thickness is carefully controlled. We show that ultra-thin silicon oxide films are of very high quality. Similar investigations are made on a twisted bonded silicon substrate obtained by the molecular bonding of two silicon wafers. It is shown that the silicon oxide is also of very good quality and can be used as a sacrificial silicon oxide in thinning down the upper silicon film. Controlled, accurate thinning is achieved down to a thickness of 10 nm, the level which is required for etching the dislocation network present at the bonding interface.     

Investigation of ferromagnetic coupling between Co layers in a Co/Pt multilayer with perpendicular anisotropy

The coercivity of a Co/Pt multilayer with out-of-plane anisotropy can be lowered greatly if it is grown onto an ultrathin NiO underlayer . By making use of this characteristic, a series of samples glass/NiO(10 Å)/[Co(4 Å)/Pt(5 Å)]₃/Pt(x Å)/[Co(4 Å)/Pt(5 Å)]₃ with different Pt spacer thickness have been prepared to determine the ferromagnetic (FM) coupling between Co layers across the Pt layer. The measurements of major and minor hysteresis loops have shown that the FM coupling between the top and bottom Co/Pt multilayers decreases monotonically with the Pt layer thickness and disappears above the Pt layer thickness of 40 Å. This thickness of 40 Å is much larger than that in the literature. In addition to the FM coupling between the top and bottom Co/Pt multilayers across the Pt spacer, there exists a weak biquadratic coupling, which induces the broad transition of the bottom Co/Pt multilayer.     

Inter-diffusion of cobalt and silicon through an ultra thin aluminum oxide layer

Optical emission spectroscopy of sputtered species during ion bombardment, Auger electron spectroscopy and high-resolution transmission electron microscopy were used to study the cobalt and silicon diffusion through the interfaces of Co/AlO/Si(0 0 1) hetero-structure. The results are discussed as a function of the annealing temperature of sample and show that the diffusion process at the interfaces starts for annealing temperatures above 200 °C without detectable modification of the oxide layer.     

Simulation and fabrication study of porous silicon photonic crystal

The present work reports design and fabrication of porous silicon based one-dimensional (1D) photonic crystal. Distributed Bragg reflector (DBR) is a 1D photonic crystal composed of multilayer stack of high and low refractive index layers. Design of porous silicon DBR is a complex one and requires appropriate control in optical parameters of its constituent layers. In order to design DBR, two porous silicon single layer samples were fabricated using current density of 10 and 50 mA/cm². Optical characterization of single layer samples showed series of interference fringes. Reflective interferometric Fourier transform spectroscopy (RIFTS) method was employed to determine optical constants of porous silicon single layers. DBR simulation was carried out based on transfer matrix method. DBR was then fabricated using optical parameters obtained from RIFTS method. Reflection bandwidth of prepared DBR was found to be 216 nm, which is comparable to the simulated value of 203 nm.     

Investigation of chemical mechanical polishing of zinc oxide thin films

Zinc oxide has become an important material for various applications. Commercially available zinc oxide single crystals and as-grown zinc oxide thin films have high surface roughness which has detrimental effects on the growth of subsequent layers and device performance. A chemical mechanical polishing (CMP) process was developed for the polishing of zinc oxide polycrystalline thin films. Highly smooth surfaces with RMS roughness <6 Å (as compared to the initial roughness of 26 ± 6 Å) were obtained under optimized conditions with removal rates as high as 670 Å/min. Effects of various CMP parameters on removal rate and surface roughness were evaluated. The role of pH on the polishing characteristics was investigated in detail.     

Effect of thickness and porous structure of SiC layers on the spectral response of the Pd/SiC-pSi Schottky photodiodes

In this work, we study the effect of the thickness and porous structure of silicon carbide (PSC) layers on the electrical properties of Schottky photodiodes by using a palladium (Pd) layer deposited on non-porous silicon carbide (SiC) and porous-SiC (PSC) layers. The non-porous and porous-SiC layers were realized on a p-type silicon (Si(1 0 0)) substrate by pulsed laser deposition using a KrF laser (248 nm) and thermal deposition of a thin Pd layer. The porous structure of the SiC layer deposited was developed by an electrochemical (anodization) method. The electrical measurements were made at room temperature (295 K) in an air ambience. The effect of the porous surface structure and the thickness of the SiC layer were investigated by evaluating electrical parameters such as the ideality factor (n) and barrier height (?_Bp). The thickness of the porous layer significantly affects the electrical properties of the Schottky photodiodes. Analysis of current-voltage (I-V) characteristics showed that the forward current might be described by a classical thermal emission theory. The ideality factor determined by the I-V characteristics was found to be dependent on the SiC thickness a value For a thin SiC layer (0.16 μm) n was around 1.325 with a barrier height 0.798 eV, while for a thick layer (1.6 μm), n and ?_Bp were 1.026 and 0.890 eV, respectively for Pd/SiC-pSi. These results indicate Schottky photodiodes with high performance are obtained for thicker SiC layer and for thin layer of PSC. This effect showed the uniformity of the SiC layer. In the same case the ideality factor (n) decreases for Pd/PSC-pSi(1 0 0) for low SiC thickness by report of Pd/PSC-pSi(1 0 0) Schottky photodiodes, but for Pd/PSC-pSi(1 0 0) n increase for large SiC thickness layer. We notice that the barrier height (?_Bp) was reversely depend by report of ideality factor. A spectral response value of (SR) of 34 mA/W at λ = 400 nm was measured for Pd/0.16 μm SiC-pSi Schottky photodiode with low SiC thickness. On the other hand, a value of SR = 0.14 mA/W at λ = 900 nm was obtained when we used PSC layer (Pd/PSC-pSi(1 0 0)). A reverse behaviour occurs for thicker SiC layer. Finally, it was found that the thickness and surface porous structure have strong effect on sensitivity.     

Structural and magnetic properties of Pt in Co/Pt multilayers

[Co(30 Å)/Pt(x Å)]₂₀ multilayers with the Pt layer thicknesses varying from 5 Å to 20 Å were characterized structurally by high angle X-ray diffraction, X-ray reflectivity, X-ray absorption spectroscopy and magnetically by X-ray magnetic circular dichroism. It is found that the structure and magnetic properties of Pt have a strong correlation with the Pt layer thickness. The 20 Å thickness Pt layer is not almost influenced by the adjacent Co layer and the nearest neighbors are dominated by Pt-Pt shells. With decreasing Pt layer thickness, the nearest neighbors are gradually dominated by Pt-Co shells and the Pt-Co intermixing regions also remarkable increase at the interfaces, especially for the 5 Å thickness Pt layer. The orbital and spin magnetic moments as well as the ratio m_orb/m_spin all decrease systematically with increasing Pt layer thickness, indicating that the interface atoms are polarized by direct Pt-Co hybridization, but that the adjacent layers are polarized by Pt-Pt interactions.     

Oxidation of Ni(Pt)Si by molecular vs. atomic oxygen

X-ray photoelectron spectroscopy (XPS) has been used to characterize the oxidation of a clean Ni(Pt)Si surface under two distinct conditions: exposure to a mixed flux of atomic and molecular oxygen (O + O₂; PO+O₂ = 5 × 10⁻⁶ Torr) and pure molecular oxygen (O₂; PO₂ = 10⁻⁵ Torr) at ambient temperatures. Formation of the clean, stoichiometric (nickel monosilicide) phase under vacuum conditions results in the formation of a surface layer enriched in PtSi. Oxidation of this surface in the presence of atomic oxygen initially results in formation of a silicon oxide overlayer. At higher exposures, kinetically limited oxidation of Pt results in Pt silicate formation. No passivation of oxygen uptake of the sample is observed for total O + O₂ exposure <8 × 10⁴ L, at which point the average oxide/silicate overlayer thickness is 23 (3) Å (uncertainty in the last digit in parentheses). In contrast, exposure of the clean Ni(Pt)Si surface to molecular oxygen only (maximum exposure: 5 × 10⁵ L) results in slow growth of a silicon oxide overlayer, without silicate formation, and eventual passivation at a total average oxide thickness of 8(1) Å, compared to a oxide average thickness of 17(2) Å (no silicate formation) for the as-received sample (i.e., exposed to ambient.) The aggressive silicon oxidation by atomic oxygen, results in Ni-rich silicide formation in the substrate and the kinetically limited oxidation of the Pt.     

Metal organic chemical vapor deposition of ultrathin ZrO2 films on Si(1 0 0) and Si(1 1 1) studied by electron spectroscopy

The growth of ultrathin ZrO₂ films on Si(1 0 0)-(2 × 1) and Si(1 1 1)-(7 × 7) has been studied with core level photoelectron spectroscopy and X-ray absorption spectroscopy. The films were deposited sequentially by chemical vapor deposition in ultra-high vacuum using zirconium tetra-tert-butoxide as precursor. Deposition of a > 50 Å thick film leads in both cases to tetragonal ZrO₂ (t-ZrO₂), whereas significant differences are found for thinner films. On Si(1 1 1)-(7 × 7) the local structure of t-ZrO₂ is not observed until a film thickness of 51 Å is reached. On Si(1 0 0)-(2 × 1) the local geometric structure of t-ZrO₂ is formed already at a film thickness of 11 Å. The higher tendency for the formation of t-ZrO₂ on Si(1 0 0) is discussed in terms of Zr-O valence electron matching to the number of dangling bonds per surface Si atom. The Zr-O hybridization within the ZrO₂ unit depends furthermore on the chemical composition of the surrounding. The precursor t-butoxy ligands undergo efficient C-O scission on Si(1 0 0), leaving carbonaceous fragments embedded in the interfacial layer. In contrast, after small deposits on Si(1 1 1) stable t-butoxy groups are found. These are consumed upon further deposition. Stable methyl and, possibly, also hydroxyl groups are found on both surfaces within a wide film thickness range.     

Signatures of epitaxial graphene grown on Si-terminated 6H-SiC (0 0 0 1)

Epitaxial graphene layers thermally grown on Si-terminated 6H-SiC (0 0 0 1) have been probed using Auger electron spectroscopy, Raman microspectroscopy, and scanning tunneling microscopy (STM). The average multilayer graphene thickness is determined by attenuation of the Si (L₂₃VV) and C (KVV) Auger electron signals. Systematic changes in the Raman spectra are observed as the film thickness increases from one to three layers. The most striking observation is a large increase in the intensity of the Raman 2D-band (overtone of the D-band and also known as the G′-band) for samples with a mean thickness of more than ∼1.5 graphene layers. Correlating this information with STM images, we show that the first graphene layer imaged by STM produces very little 2D intensity, but the second imaged layer shows a single-Lorentzian 2D peak near 2750 cm⁻¹, similar to spectra acquired from single-layer micromechanically cleaved graphene (CG). The 4-10 cm⁻¹ higher frequency shift of the G peak relative to CG can be associated with charge exchange with the underlying SiC substrate and the formation of finite size domains of graphene. The much greater (41-50 cm⁻¹) blue shift observed for the 2D-band may be correlated with these domains and compressive strain.     

Investigation of the interaction between electrical discharges and low resistivity silicon substrates

In this work the impact of single discharge pulses in air on single-crystalline, p-type silicon having a low bulk resistivity of 0.009-0.012 Ω cm is investigated. Compared to platinum specimens, the craters in silicon have lateral dimensions which are about one order of magnitude larger despite comparable values for the melting point and the melting energy. This finding is attributed to the substantially higher bulk resistivity of silicon leading a higher energy input into the substrate when spark loaded. The energy generated by joule heating is, however, distributed across a larger area due to a current spreading effect. To study the impact of different surface properties on the sparking behaviour, the crater formation on the silicon substrate is investigated applying coatings with different material properties, such as sputter-deposited aluminium layers and thermally-grown silicon dioxide. In general, the crater characteristics formed on unmodified silicon is not influenced when a thin aluminium layer of 24 nm is deposited. At higher film thickness above 170 nm, the sparking energy is almost completely absorbed in the top layer with low influence on the underlying silicon substrate. In the case of a dielectric top layer with a thickness of 155 nm, the formation of many small distinct craters is supported in contrast to a 500 nm-thick SiO₂ film layer where the generation of a single crater with a large area is energetically favoured. A surface roughness of several nm on the silicon probes has no measurable effect on crater formation when compared to an original surface characteristic with values in the sub-nm range.