Ion beam as a probe to study the behavior of hydrogen on silicon surfaces

Hydrogen adsorption and its behavior on Si surfaces is studied by ion beam techniques in the energy range of MeV–keV. Elastic recoil detection analysis employing MeV ion beams is one of the most reliable experimental techniques for direct determination of absolute hydrogen coverages on Si surfaces. Results of its application to Si(1 0 0) and Si(1 1 1) clean surfaces are described. Important new results of the role of adsorbed hydrogen on the growth process or structures of metallic thin films on Si(1 1 1) surfaces also are described. Characterization of the growth process or structure of the thin films, as well as the characterization of hydrogen, is performed by ion beam techniques.     

Adsorption of water monomer and clusters on platinum(111) terrace and related steps and kinks: I. Configurations,energies, and hydrogen bonding

Adsorption and rotation of water monomer, dimer, and trimer on the (111) terrace, (221) and (322) stepped, and (763) and (854) kinked surfaces of platinum were studied by density functional theory calculations using the PW91 approximation to the energy functional. On the (111) terrace, water monomer and the donor molecule of the dimer and trimer adsorb at atop sites. The per-molecule adsorption energies of the monomer, dimer, and trimer are 0.30, 0.45, and 0.48 eV, respectively. Rotation of monomers, dimers, and trimers on the terrace is facile with energy barriers of 0.02 eV or less. Adsorption on steps and kinks is stronger than on the terrace, as evidenced by monomer adsorption energies of 0.46 to 0.55 eV. On the (221) stepped surface the zigzag extended configuration is most stable with a per-molecule adsorption energy of 0.57 eV. On the (322) stepped surface the dimer, two configurations of the trimer, and the zigzag configuration have similar adsorption energies of 0.55 ± 0.02 eV. Hydrogen bonding is strongest in the dimer and trimer adsorbed on the terrace, with respective energies of 0.30 and 0.27 eV, and accounts for their increased adsorption energies relative to the monomer. Hydrogen bonding is weak to moderate for adsorption at steps, with energies of 0.04 to 0.15 eV, as the much stronger water–metal interactions inhibit adsorption geometries favorable to hydrogen bonding. Correlations of hydrogen bond angles and energies with hydrogen bond lengths are presented. On the basis of these DFT/PW91 results, a model for water cluster formation on the Pt(111) surface can be formulated where kink sites nucleate chains along the top of step edges, consistent with the experimental findings of Morgenstern et al., Phys. Rev. Lett., 77 (1996) 703.     

Highly sensitive hydrogen detection by medium energy Ne+ impact

Hydrogen atoms on solid surfaces were measured directly by elastic recoil detection analysis (ERDA) using medium energy (100–150 keV) Ne⁺ ions with an excellent sensitivity of (～ 1 × 10¹² H/cm²) without any absorber foils and time-of-flight techniques. An electrostatic toroidal analyzer acquired H⁺ ions with energy around 11 keV recoiled from Si(111)-1 × 1-H surfaces. The H⁺ fraction strongly depends upon emerging angle and takes a value more than 50% at the angle below 70° and a saturated value of 17% at the angle above 80° with respect to surface normal. We detected H atoms on the reduced TiO₂(110) exposed to water molecules at room temperature (2 L) and estimated the absolute amount of H to be ～ 2.0 × 10¹⁴ H/cm² corresponding to ～ 38% (～ 0.38 ML) of the bridging oxygen atoms.     

High-resolution depth profiling of ultrashallow boron implants in silicon using high-resolution RBS

Depth profiles of ultralow energy (0.2–0.5 keV) B ion implants in Si(0 0 1) samples are measured by high-resolution Rutherford backscattering spectroscopy. The boron profile does not show a narrow surface concentration peak which is usually observed in the measurement of secondary ion mass spectroscopy. The obtained boron profiles roughly agree with TRIM simulation even at 0.2-keV B ion implantation.     

Medium energy ion scattering investigation of methylthiolate-induced modification of the Au(111) surface

100 keV H⁺ scattering has been used to investigate the structure of the methylthiolate/Au(111) interface in the Au(111)(√3 × √3)R30° phase. Adsorption of the thiolate onto the clean Au(111) surface leads to a large drop in the scattered ion yield due to the lifting of the clean surface ‘herring-bone’ reconstruction, but the thiolate-covered surface shows an ion yield higher than that of an unreconstructed Au(111) surface, providing direct evidence of a significant number of Au atoms that are displaced from their bulk-terminated positions at the buried interface. Simulations for two different Au adatoms models at the interface, namely, the Au-adatom-monothiolate (AAM) and Au-adatom-dithiolate (AAD) models, show significant sensitivity to the exact values of interlayer spacings and atomic vibrational amplitudes, but the comparison with experimental results appears to favour the AAD model with 0.17 ML Au adatoms in bridging sites at the interface.     

Low energy alkali ion scattering investigation of Au nanoclusters grown on silicon oxide surfaces

The atomic and electronic structures of Au nanostructures grown by deposition onto various silicon oxide surfaces were probed with low energy alkali ion scattering. Charge state-resolved time-of-flight spectra of scattered 2 keV ³⁹K⁺ ions were collected from Au deposited onto an untreated Si wafer with a native oxide, a thermally grown oxide surface, and atomically-clean Si(111). It is shown that nanoclusters form on both oxides, but not on the clean Si. A quantitative analysis of the ion scattering spectra indicates that the nanoclusters are initially flat, two-dimensional structures that start to develop a second layer at about 0.5 Å of deposited Au and then form three-dimensional islands. The neutral fraction of scattered 2 keV ³⁹K⁺ ions decreases with deposition indicating changes in the quantum state occupancy with cluster size. The shapes of the clusters differ on the native and thermal oxides, leading to shape-dependent neutralization.     

Attenuation lengths of high energy photoelectrons in compact and mesoporous SiO2 films

We have experimentally evaluated attenuation lengths (AL) of photoelectrons traveling in compact and micro and mesoporous (~ 45% voids) SiO₂ thin films with high (8.2–13.2 keV) kinetic energies. The films were grown on polished Si(100) wafers. ALs were deduced from the intensity ratio of the Si 1s signal from the SiO₂ film and Si substrate using the two-peaks overlayer method. We obtain ALs of 15–22 nm and 23–32 nm for the compact and porous SiO₂ films for the range of kinetic energies considered. The observed AL values follow a power law dependence on the kinetic energy of the electrons where the exponent takes the values 0.81 ± 0.13 and 0.72 ± 0.12 for compact and porous materials, respectively.     

Interaction of HNCO with Au(111) surfaces

The surface chemistry of isocyanic acid, HNCO, and its dissociation product, NCO, was studied on clean, O-dosed and Ar ion bombarded Au(111) surfaces. The techniques used are high resolution energy loss spectroscopy (HREELS) and temperature-programmed desorption (TPD). The structure of Ar ion etched surface is explored by scanning tunneling microscopy (STM). HNCO adsorbs molecularly on Au(111) surface at 100 K yielding strong losses at 1390, 2270 and 3230 cm^? 1. The weakly adsorbed HNCO desorbs in two peaks characterized by T_p = 130 and 145 K. The dissociation of the chemisorbed HNCO occurs at 150 K to give NCO species characterized by a vibration at 2185 cm^? 1. The dissociation process is facilitated by the presence of preadsorbed O and by defect sites on Au(111) produced by Ar ion bombardment. In the latter case the loss feature of NCO appeared at 2130 cm^? 1. Isocyanate on Au(111) surface was found to be more stable than on the single crystal surfaces of Pt-group metals. Results are compared with those obtained on supported Au catalysts.     

Structural dependence of intermediate species for the hydrogen evolution reaction on single crystal electrodes of Pt

Adsorbed hydrogen and water were measured during the hydrogen evolution reaction (HER) on the low and high index planes of Pt in 0.5 M H₂SO₄ using infrared reflection absorption spectroscopy. Hydrogen is adsorbed at the atop site (atop H) on Pt(110) during the HER, whereas adsorbed hydrogen at the asymmetric bridge site (bridge H) is found on Pt(100). The band intensity of the adsorbed hydrogen depends on temperature, indicating that the bands are due to the intermediate species for the HER. The band of the atop H appears on stepped surfaces with (110) step, whereas the asymmetric bridge H is observed on Pt(211) = 3(111)–(100) and Pt(311) = 2(111)–(100) that have (100) step. The absence of the atop H on Pt(100), Pt(211), and Pt(311) can be attributed to the relative stability of the bridge site.     

Ab initio study of Yb on the Ge(111)–(3 × 2) and Si(111)–(3 × 2) surfaces

The surface reconstruction, 3 × 2, induced by Yb adsorption on a Ge (Si)(111) surface has been studied using first principles density-functional calculation within the generalized gradient approximation. The two different possible adsorption sites have been considered: (i) H₃ (this site is directly above a fourth-layer Ge (Si) atom) and (ii) T₄ (directly above a second-layer Ge (Si) atom). We have found that the total energies corresponding to these binding sites are nearly the same, indeed for the Yb/Ge (Si)(111)–(3 × 2) structure the T₄ model is slightly energetic by about 0.01 (0.08) eV/unitcell compared with the H₃ model. In particular for the Ge sublayer, the energy difference is small, and therefore it is possible that the T₄, H₃, or T₄H₃ (half of the adatoms occupy the T₄ adsorption site and the rest of the adatoms are located at the H₃ site) binding sites can coexist with REM/Ge(111)–(3 × 2). In contrast to the proposed model, we have not determined any buckling in the Ge = Ge double bond. The electronic band structures of the surfaces and the corresponding natures of their orbitals have also been calculated. Our results for both substrates are seen to be in agreement with the recent experimental data, especially that of the Yb/Si(111)–(3 × 2) surface.     

Structural analysis of Si(111)-√21 × √21-(Ag,Cs) surface by reflection high-energy positron diffraction

We have investigated the Si(111)-√21 × √21-(Ag, Cs) superstructure using reflection high-energy positron diffraction. Rocking curve analysis based on the dynamical diffraction theory reveals that Cs atoms are located at a height of 3.04 Å above the underlying √3 × √3-Ag structure and that they form a triangular structure with a side length of 10.12 Å. The structure of the Si(111)-√21 × √21-(Ag, Cs) surface is significantly different from those of the Si(111)-√21 × √21-Ag and Si(111)-√21 × √21-(Ag, Au) surfaces, probably because of the different electronic structures of the alkali and noble metal atoms.     

Interfacial structure of Co porphyrins on Au(111) electrode: Interaction of porphyrin molecules with substrate

Interfacial structures of cobalt(II) porphine (CoP) and [2,3,7,8,12,13,17,18‐octaethyl‐21H,23-H-porphine]cobalt(II) (CoOEP) have been studied on Au(111) electrode using electrochemical scanning tunneling microscopy (EC-STM), in-situ X-ray diffraction, and density functional theory (DFT) calculations. The adsorption of porphyrins affects the reconstruction of Au(111) surface. The adsorption of CoP causes a lifting of the reconstruction to a complete 1 × 1 structure of Au(111). On CoOEP modified Au(111), the unit cell periodicity of the reconstructed substrate structure expands compared with the √3 × 23 structure of bare Au(111). The same expanded substrate structure was observed on Au(111) modified with OEP without the coordinated Co ion; the coordinated metal ion of the adsorbed porphyrin molecule does not affect the substrate structure. This result indicates that the interaction of conjugated π electrons of porphyrin with the substrate is stronger than that of the coordinated Co ion. In-situ X-ray diffraction and DFT calculation support non-covalent interaction of porphyrins with the Au(111) surface.     

CO adsorption on clean and oxidized Pt3Ti(111)

CO adsorption on clean and oxidized Pt₃Ti(111) surfaces has been investigated by means of Auger Electron Spectroscopy (AES), Thermal Desorption Spectroscopy (TDS), Low Energy Electron Diffraction (LEED) and High Resolution Electron Energy Loss Spectroscopy (HREELS). On clean Pt₃Ti(111) the LEED patterns after CO adsorption exhibit either a diffuse or a sharp c(4 × 2) structure (stable up to 300 K) depending on the adsorption temperature. Remarkably, the adsorption/desorption behavior of CO on clean Pt₃Ti(111) is similar to that on Pt(111) except that partial CO decomposition on Ti sites and partial CO oxidation have also been evidenced. Therefore, the clean surface cannot be terminated by a pure Pt plane. Partially oxidized Pt₃Ti(111) surfaces (< 135 L O₂ exposure at 1000 K) exhibit a CO adsorption/desorption behavior rather similar to that of the clean surface, showing again a c(4 × 2) structure (stable up to 250 K). Only the oxidation of CO is not detectable any more. These results indicate that some areas of the substrate remain non-oxidized upon low oxygen exposures. Heavily oxidized Pt₃Ti(111) surfaces (> 220 L O₂ exposure at 1000 K) allow no CO adsorption indicating that the titanium oxide film prepared under these conditions is completely closed.     

Methylamine decomposition on nickel surfaces: A density functional theory study

The adsorption and decomposition of methylamine on Ni(1 1 1), Ni(1 0 0), stepped Ni(1 1 1), and nitrogen atom modified Ni(1 0 0) (denoted N–Ni(1 0 0)) have been studied with the DFT–GGA method using the periodic slab models. The initial scissions of C–H, N–H and C–N bond are considered. The adsorption energies under the most stable configurations for the possible species and the activation energies for the possible initial elementary reactions involved are obtained in the present work. Through systematic exploring of the kinetics mechanism of methylamine decomposition on these four surfaces, it is found that the reactivity of these surfaces decreased with the order of stepped Ni(1 1 1) > Ni(1 0 0) > Ni(1 1 1) > N–Ni(1 0 0). This indicates that the reactivity is related to the openness of the surface, and the presence of nitrogen atom reduces the reactivity of the Ni(1 0 0). For the three reactions, the barriers decreased with the order of C–N > N–H > C–H on Ni(1 1 1) and Ni(1 0 0), whereas they decreased with the order of C–N > C–H > N–H on stepped Ni(1 1 1) and N–Ni(1 0 0).     

Mechanism of negative ion emission from surfaces of ferroelectrics

Experimental studies of the mechanism of negative ion and cluster ion emission from surfaces of ferroelectrics are described. The emission was produced by negative voltage pulses with the amplitude of about 400 V, with a rapid rise-time (below 10 ns) and a slow decay‐time (several μs). The pulses were applied between the back side of the ferroelectric sample and the metal tip touching the front emitting side. The surface of the ferroelectrics could be cleaned in situ by 1 keV Ar⁺ bombardment. The morphologic changes around the tip were observed with an atomic force microscope (AFM). Mostly negative ions and cluster ions were emitted and studied in our experiments. Positive ions were detected with much lower probability and are produced by an entirely different microscopic process than negative ions. Masses as well as energies of emitted ions were measured with a time-of-flight (TOF) spectrometer and compared with available spontaneous desorption (SD) spectra and Cs‐SIMS spectra in order to clarify the mechanism of the emission. The trajectory of ions emitted from the sample was studied by computer simulation. The conclusion of these studies is that the negative ion emission is caused by the Coulomb explosion of a polarization cloud rapidly formed at the front edge of the pulse. The explosion takes place in the vicinity of the tip-sample contact at distances several tens of μm from the contact where the stabilizing effect of the positively charged tip is already small.     

Application of low energy ion blocking for adsorption site determination of Na Atoms on a Cu(111) surface

Ion blocking in the low keV energy range is demonstrated to be a sensitive method for probing surface adsorption sites by means of the technique of time-of-flight scattering and recoiling spectroscopy (TOF-SARS). Adsorbed atoms can block the nearly isotropic backscattering of primary ions from surface atoms in the outmost layers of a crystal. The relative adsorption site position can be derived unambiguously by simple geometrical constructs between the adsorbed atom site and the surface atom sites. Classical ion trajectory simulations using the scattering and recoiling imaging code (SARIC) and molecular dynamics (MD) simulations provide the detailed ion trajectories. Herein we present a quantitative analysis of the blocking effects produced by sub-monolayer Na adsorbed on a Cu(111) surface at room temperature. The results show that the Na adsorption site preferences are different at different Na coverages. At a coverage θ = 0.25 monolayer, Na atoms preferentially populate the fcc threefold surface sites with a height of 2.7 ± 0.1 Å above the 1st layer Cu atoms. At a lower coverage of θ = 0.10 monolayer, there is no adsorption site preference for the Na atoms on the Cu(111) surface.     

Structure and dynamics of octamethyl-ethinyl-ferrocene: an organometallic rotator phase

The onset of dynamics and the structural changes of octamethyl–ethinyl-ferrocene around a solid–solid phase transition at 248 K have been investigated using the novel technique of quasielastic nuclear forward scattering (QNFS) in a temperature range between 61 and 257 K at a photon energy of 14.413 keV, as well as X-ray powder diffraction at various energies. A pronounced hysteresis for both the dynamical and the structural properties is observed and confirmed by differential-scanning calorimetry. We assign the observed phenomena to a first-order transition from a low-symmetry low-temperature phase to a high-symmetry high-temperature plastically crystalline phase with nearly cubic symmetry. Possible mechanisms for this transition are discussed in the light of our results.     

Sputter damage in Si surface by low energy Ar+ ion bombardment

The damage distributions in Si(1 0 0) surface after 1.0 and 0.5 keV Ar⁺ ion bombardment were studied using MEIS and Molecular dynamic (MD) simulation. The primary Ar⁺ ion beam direction was varied from surface normal to glancing angle. The MEIS results show that the damage thickness in 1.0 keV Ar ion bombardment is reduced from about 7.7 nm at surface normal incidence to 1.3 nm at the incident angle of 80°. However, the damage thickness in 0.5 keV Ar ion bombardment is reduced from 5.1 nm at surface normal incidence to 0.5 nm at the incident angle of 80°. The maximum atomic concentration of implanted Ar atoms after 1 keV ion bombardment is about 10.5 at% at the depth of 2.5 nm at surface normal incidence and about 2.0 at% at the depth of 1.2 nm at the incident angle of 80°. However, after 0.5 keV ion bombardments, it is 8.0 at% at the depth of 2.0 nm for surface normal incidence and the in-depth Ar distribution cannot be observable at the incident angle of 80°. MD simulation reproduced the damage distribution quantitatively.     

TiO2 chemical vapor deposition on Si(111) in ultrahigh vacuum: Transition from interfacial phase to crystalline phase in the reaction limited regime

The interaction between the metal organic precursor molecule titanium(IV) isopropoxide (TTIP) and three different surfaces has been studied: Si(111)-(7 × 7), SiO_x/Si(111) and TiO₂. These surfaces represent the different surface compositions encountered during TTIP mediated TiO₂ chemical vapor deposition on Si(111). The surface chemistry of the titanium(IV) isopropoxide precursor and the film growth have been explored by core level photoelectron spectroscopy and x-ray absorption spectroscopy using synchrotron radiation. The resulting film morphology has been imaged with scanning tunneling microscopy. The growth rate depends on both surface temperature and surface composition. The behavior can be rationalized in terms of the surface stability of isopropoxy and isopropyl groups, confirming that growth at 573 K is a reaction limited process.     

Dynamical LEED analyses of the clean and the NO-adsorbed Ir(111) surface

The adsorption structure of nitric oxide (NO) on Ir(111) was studied by thermal desorption spectroscopy (TDS) and dynamical analyses of low-energy electron diffraction (LEED). At the saturation coverage at about 100 K, a 2 × 2 pattern was observed by LEED and two peaks appeared at 365 and 415 K in TDS. No change in the LEED I–V curves was observed by annealing at 280 K, which means that the NO-saturated surface was retained at this temperature. On the contrary, partial desorption and changes of the LEED I–V curves were observed by annealing at 360 K. Combined with previous vibrational studies, it is suggested that one adsorption species is not affected, while another species is partially desorbed and the rest of them are dissociated by annealing at 360 K. Dynamical analyses of LEED were performed for the 280 K-annealed and the 360 K-annealed surfaces, which correspond to the NO-saturated and the NO-dissociated Ir(111) surfaces, respectively. These revealed that NO occupies the atop, fcc-hollow and hcp-hollow sites (atop-NO + fcc-NO + hcp-NO) for the NO-saturated Ir(111) surface with the saturation coverage of 0.75 ML. For the 360 K-annealed surface, the atop-NO is not affected but the fcc-NO and the hcp-NO are partially desorbed as NO and partially dissociated to N and O, both of which occupy the fcc-hollow site on the surface.