Search

Nanoscale electron field emitters are known to produce more stable electron emissions than conventional emitters. This has been attributed to size effects; nanoscale emitters can operate with a small emission current and a low extraction voltage, which reduces the bombardment of residual gas ions on the emitter tip. However, our experiments discovered that nanoscale LaB

A material with a low work function exhibiting field-emission of electrons has long been sought as an ideal point electron source to generate a coherent electron beam with high brightness, long service life, low energy spread, and especially stable emission current. The quality and performance of the electron source are now becoming limiting factors for further improving the spatial resolution and analytical capabilities of the electron microscope. While tungsten (W) is still the only material of choice as a practically usable field emission filament since it was identified more than six decades ago, its electron optical performance remains unsatisfactory, especially the poor emission stability (>5% per hour), rapid current decay (20% in 10 hours), and relatively large energy spread (0.4 eV), even in an extremely high vacuum (10−9 Pa). Herein, we report a LaB6 nanoneedle structure having a sharpened tip apex with a radius of curvature of about 10 nm that is fabricated and finished using a focused ion beam (FIB) and show that it can produce a field emission electron beam meeting the application criteria with a high reduced brightness (1010 A m−2 sr−1 V−1), small energy spread (0.2 eV), and especially high emission stability (

Electric field-induced point electron source is highly demanded for microscopy, spectroscopy, lithography, X-ray tubes, microwave devices, and data displays. However, the instability in emission current and requirement of ultrahigh vacuum have often limited its extensive applications. Herewith we report a single-crystalline HfC nanowire with oxycarbide emission surface for stable electron emission at 50 nA with fluctuations less than 1% in a vacuum of 4 × 10−7 Pa. The emitter has a low work function of 2.5 eV measured by the field emission Fowler-Nordheim curve and it is in good agreement with density functional theory (DFT) calculations. The energy spread is in a range of 0.21–0.26 eV with a corresponding reduced brightness 1.95 × 1011−3.81 × 1011 A·m−2·sr−1·V−1. The HfC nanowire with oxycarbide emission surface is a qualified candidate for the next-generation electron source with high brightness, large current, and low energy spread.

Phonon transport across a vacuum gap separating intrinsic silicon crystals is predicted via the atomistic Green's function method combined with first-principles calculations of all interatomic force constants. The overlap of electron wave functions in the vacuum gap generates weak covalent interaction between the silicon surfaces, thus creating a pathway for phonons. Phonon transport, dominated by acoustic modes, exceeds near-field radiation for vacuum gaps smaller than ~ 1 nm. The first-principles-based approach proposed in this work is critical to accurately quantify the contribution of phonon transport to heat transfer in the extreme near field., 31 pages, 4 figures, 4 supplemental figures

Searching for viable field-emission structures with a stable emission current, high emission brightness, and low energy spread has always been highly demanded for the development of high-performance electron-optical instrumentation. In this work, a single-crystalline HfC nanowire emitter with an oxycarbide surface is investigated. This new emitter shows an excellent current stability with a fluctuation less than 1%/h, as well as a narrow energy spread and high emission brightness. It exhibits great potentials to be a practical electron source for experimental trials in electron microscopes.

Graphite-like BC2 has yet to be synthesized; however, it stably exists in the Sc2B1.1C3.2 compound, where the boron atoms are arranged as far apart from each other as possible. Recently, a theoretical study on monolayer BC2 reported that in the most stable structure, B atoms are positioned adjacent to each other. We anticipated that graphite-like BC2 might take a different structure based on the electric charge. Therefore, we carried out first principles calculations to investigate whether this is true or not. The most stable structure among the six possible structures changed with the increase in the negative electric charge, which well explained both the previous results without contradiction. The most stable structure was also dependent on the pressure. The Li intercalation potential for BC2 was calculated to investigate its applicability as an anode for lithium-ion batteries. Our results revealed that Li atoms can be intercalated into BC2 to yield Li1.5BC2, whose gravimetric capacity is approximately 3.1 times higher than that of LiC6. However, the most stable structure with intercalated Li atoms became unstable when all the Li atoms were extracted. This feature may hinder the repetitive charge–discharge cycle of the anode and hence needs to be carefully considered.

Herein, the work functions of group 4 and group 5 transition-metal (Ti, Zr, Hf, V, Nb, and Ta) carbides and transition-metal oxycarbides (TMCOs) were investigated by first-principles calculations for their potential application as electron emitters. The work functions of both groups decreased as the substitution of carbon atoms with oxygen proceeded, and the reduction in group 4 was more than that of group 5. In particular, ZrC1-x O x and HfC1-x O x (x ≥ 0.25) exhibited work functions of less than 3 eV, which were comparable with those of LaB6- and ZrO-coated tungsten. The reduction in the work functions could be explained by the rigid-band model of the electronic density of states. The increase in valence electrons increased the Fermi energy, while it demonstrated a less significant influence on the vacuum potential, resulting in a reduction in the work functions. The phonon dispersion curves indicated that the NaCl-type group 5 TMCOs were less stable than the group 4 TMCOs. This agrees with the experimental findings that TaC1-x O x was not synthesized and NbC1-x O x was synthesized only for smaller values of x (i.e., x < 0.28). From the viewpoints of the work functions and structural stabilities, group 4 (Ti, Zr, and Hf) TMCOs exhibit better potential for application as electron emitters than group 5 (V, Nb, and Ta) TMCOs.

A material with a low work function exhibiting field-emission of electrons has long been sought as an ideal point electron source to generate a coherent electron beam with high brightness, long service life, low energy spread, and especially stable emission current. The quality and performance of the electron source are now becoming limiting factors for further improving the spatial resolution and analytical capabilities of the electron microscope. While tungsten (W) is still the only material of choice as a practically usable field emission filament since it was identified more than six decades ago, its electron optical performance remains unsatisfactory, especially the poor emission stability (5% per hour), rapid current decay (20% in 10 hours), and relatively large energy spread (0.4 eV), even in an extremely high vacuum (10

The ternary borocarbide Sc2B1.1C3.2 is known to have a layered structure in which graphite-like BC2 layers and NaCl-like Sc2C layers are alternately stacked. Due to the complexity of the structure, X-ray analysis is unable to determine the precise structure of the BC2 layer, namely, how the boron atoms are arranged in the layer. Furthermore, the stacking manner of the two layers remains unclear. In this study, we have investigated the entire structure of Sc2B1.1C3.2 through first-principles calculations. The calculations reveal that the material is the most stable when B atoms are arranged to form 3 × 3 –R30° superlattices in the graphite-like layer. The interlayer bonding is mostly van der Waals-like, and the total energy is almost independent of the stacking method. The BC2 layer with the 3 × 3 –R30° structure can be regarded as an intermediate between similar graphite-like materials, BC and BC3. The electronic density of states of Sc2B1.1C3.2 exhibits metallic features, and electrons are transferred from Sc to B and C atoms, which compensates the lack of electrons in the BC2 layer. The origin of the undulation in the BC2 layer is not the inner electronic structure but the interaction between the BC2 and Sc2C layers. The weak localization found in the in-plane resistivity in experiments is explained by the domain boundaries of the 3 × 3 –R30° structure rather than the randomness of the arrangement of B atoms.

A first-order polyamorphic (amorphous–amorphous) phase transition is often observed in amorphous materials that have a tetrahedral structure, such as ice, carbon, or silicon, but it has yet to be observed in amorphous boron nitride (a-BN). We investigated the existence of a first-order phase transition in a-BN by means of first-principles molecular dynamics (MD) simulations at a constant temperature and various pressures. At 300 K, the volume of a-BN gradually decreases at applied pressures of 0 to 18 GPa; it then rapidly decreases by about 3% at pressures of 18 to 24 GPa, providing evidence of a first-order phase transition. The structure remains amorphous throughout the transition. As the pressure is released from 24 to 0 GPa, the volume gradually increases with preservation of the difference at the phase transition, showing that the structure of the high-density phase is maintained at 0 GPa. An analysis of the coordination number revealed that a-BN consists mainly of sp 2 -hybridized B N bonds at 0 GPa. At the phase transition, sp 2 bonds are rapidly converted into sp 3 bonds, which subsequently account for about 30% of all bonds. This mechanism resembles that of the phase transition of amorphous carbon, in which the conversion rate is almost 100%. The low conversion rate in a-BN is probably attributable to the relative inflexibility of the structure.

The field-emission characteristics of zirconium carbide (ZrC) nanowire with an oxidized surface are studied. When the ZrC nanowire is processed to introduce an oxidized surface, its field-emission performance is improved significantly, exhibiting a high field-emission current and enhanced emission current stability with a fluctuation of 1.7% in a vacuum of 6 × 10−8 Pa. Furthermore, a reduced turn-on voltage of the ZrC nanowire after oxidation was observed in the field-emission measurement, which is attributed to a reduction of the work function of the ZrC nanowire after oxidation treatment. Density functional theory calculations have also been performed to validate the mechanisms responsible for the improvement of field-emission stability and reduction in the work function of the ZrC nanowire.

This review discusses the ternary borocarbide Sc2B1.1C3.2 and graphite-like (g-) BC2. Sc2B1.1C3.2 has a layered structure in which two types of layers, graphene-like BC2 and MXene-like Sc2C, are stacked alternately; hence, Sc2B1.1C3.2 can be regarded as a graphene-MXene complex compound. The X-ray diffraction structural analysis was ambiguous due to the close atomic numbers of B and C; however, the most stable structure determined by first-principles calculations had B atoms in the BC2 layer arranged in a √ 3 × √ 3 –R30° superlattice which was an intermediate between the already known BC3 and BC structures. Bulk g-BC2 has also been studied for potential application as the anode of Li-ion batteries. Calculation of the intercalation potential revealed that Li atoms can be intercalated into BC2 to yield Li1.5BC2, which has a specific capacity approximately 3.1 times that of graphite (LiC6).

It is known that elemental boron has five polymorphs: α- and β-rhombohedral, α- and β-tetragonal, and the high-pressure γ phase. β-tetragonal (β-t) boron was first discovered in 1960, but there have been only a few studies since then. We have thoroughly investigated, using first-principles calculations, the atomic and electronic structures of β-t boron, the details of which were not known previously. The difficulty of calculation arises from the fact that β-t boron has a large unit cell that contains between 184 and 196 atoms, with 12 partially-occupied interstitial sites. This makes the number of configurations of interstitial atoms too great to calculate them all. By introducing assumptions based on symmetry and preliminary calculations, the number of configurations to calculate can be greatly reduced. It was eventually found that β-t boron has the lowest total energy, with 192 atoms (8 interstitial atoms) in an orthorhombic lattice. The total energy per atom was between those of α- and β-rhombohedral boron. Another tetragonal structure with 192 atoms was found to have a very close energy. The valence bands were fully filled and the gaps were about 1.16 to 1.54 eV, making it comparable to that of β-rhombohedral boron.

Recently, a lanthanum hexaboride (LaB6) thin film exhibiting the low work function has been developed with radio frequency (RF) sputtering deposition. Here we report an experimental study on the electronic structure of a 20 nm-thick N-doped LaB6 thin film by scanning tunneling spectroscopy (STS). The obtained dI/dV spectra have been evaluated by comparison with the theoretical calculation performed on the pristine bulk LaB6. We found, even after air exposure, annealing at 500 °C revived the local electronic structure consistent with that of the pristine LaB6. Our data clearly suggest that the nitrogen-doping makes the surface chemically inert without considerably modifying the original electronic structure of LaB6. We anticipate that the N-doped LaB6 thin film enables to extend the application scope of the LaB6.

The structural stability of cagelike boron clusters with octahedral and tetrahedral symmetries has been investigated by means of first-principles calculations. Twenty-eight cluster models, ranging from B(10) to B(66), were systematically constructed using regular and semiregular polyhedra as prototypes. The binding energies per atom were, on the whole, slightly lower than those of icosahedral clusters B(80) and B(100), which are supposed to be the most stable in the icosahedral group. The larger clusters did not always have higher binding energies. Isothermal molecular dynamics simulations were performed to determine the deformation temperatures at which clusters began to break or change their structures. We found eight clusters that had nonzero deformation temperatures, indicating that they are in metastable states. The octahedral cluster B(18) had the highest deformation temperature among these, similar to that of icosahedral B(80) and B(100). The analysis of the electronic structure of B(18) showed that it attained this high stability owing to Jahn-Teller distortion.

The crystal and electronic structures of α-tetragonal (α-t) boron were investigated by first-principles calculation. Application of a simple model assuming 50 atoms in the unit cell indicated that α-t boron had a metallic density of state, thus contradicting the experimental fact that it is a p-type semiconductor. The presence of an additional two interstitial boron atoms at the 4c site made α-t boron semiconductive and the most stable. The cohesive energy per atom was as high as those of α- and β-rhombohedral boron, suggesting that α-t boron is produced more easily than was previously thought. The experimentally obtained α-t boron in nanobelt form had about two interstitial atoms at the 8i sites. We consider that the shallow potential at 8i sites generates low-energy phonon modes, which increase the entropy and consequently decrease the free energy at high temperatures. Calculation of the electronic band structure showed that the highest valence band had a larger dispersion from Γ to Z than from Γ to X; this indicated a strong anisotropy in hole conduction.

α-Rhombohedral (α-rh) boron, which is the most stable of boron's polymorphs at low temperatures, has p-type semiconductive properties. There have been some attempts to dope the interstitial sites with alkali atoms to create metallic or n-type semiconductive α-rh boron, but this has yet to be achieved. In a previous work, we proposed the codoping of α-rh boron with Li and P or As, and revealed from first principles calculations that B12PLi and B12AsLi could be synthesized and become narrow-gap semiconductors. The band structure suggested that the mobility of electrons might be greater than that of holes. In this paper, based on these prospective results, we selected a new combination of dopants, Li and O, and theoretically studied such compounds as B12OLi and B12O2Li. The results showed that both of these materials are metallic, while the reaction energies of the Li insertion into B12O and B12O2 are lower (more unstable) than with B12PLi and B12AsLi. It was proved that the differences in the electronic structures are caused by the dangling bonds of the dopant atoms, O, P and As.

The electronic structure of liquid boron and cluster formation in expanded liquid boron have been investigated with first-principles molecular dynamics simulations. The calculated electronic density of states (DOS) exhibits a metallic feature, while liquid boron is known experimentally to be semiconductive. Since the DOS is not very sensitive to density, the electronic states near the Fermi level will consist mainly of dangling bonds, which explains the difference between the calculated and experimental results. Many types of clusters are formed in expanded liquid boron. This formation occurs in a very different way from that at low temperatures because expanded liquid boron has a high temperature and pressure that are close to the liquid-gas critical point. As the density is reduced, the coordination number in boron clusters decreases to about 2, indicating that the cluster geometry tends to be one- rather than two-dimensional, which is the most stable form at low temperatures. In fact, the analysis of small clusters proved that one-dimensional forms are dominant over two- and three-dimensional forms. This is because one-dimensional geometries have a more flexible structure and a high entropy value that consequently reduces the free energy at high temperatures.

α-Rhombohedral (α-rh) boron is the most stable polymorph of elemental boron at low temperatures and exhibits p-type semiconductive properties. Although α-rh boron has some voids of sufficient size in its structure, it does not accept alkali metals as donor dopants. An n-type or metallic α-rh boronlike material so far has not been found. We considered substituting Li for either of P or As in the stable α-rh boronlike compounds B12P2 and B12As2 and calculated the electronic structures from first principles to investigate the possibility that α-rh boron becomes an n-type semiconductor or a metal. The idea is similar to the case of diamond, where the codoping of a donor and acceptor creates an n-type semiconductor. Our results showed that B12PLi and B12AsLi became intrinsic narrow-gap semiconductors with band gaps of 0.478 and 0.536 eV, respectively. In both materials, the curvature of the conduction band was higher than that of the valence band, suggesting that the electron mobility may be higher than the ho...

The surface energy of β-tetragonal (β-t) boron has been calculated from first principles to investigate how it affects nanoscale crystal growth. The results show that the surface energy of β-t boron is higher than that of α-tetragonal (α-t) boron, and lower than that of α-rhombohedral (α-rh) and β-rhombohedral (β-rh) borons. Although the cohesive energy of β-t boron is lower (more unstable) than that of α-rh and β-rh boron, β-t boron has the lowest total energy of all boron polymorphs in a region with a crystal size N = 216−512 (N: number of atoms). This explains why β-t boron can be obtained by gas-phase synthesis such as chemical vapor deposition (CVD) but not by liquid-phase synthesis because an appropriate number of atoms must be isolated to obtain the β-t structure. The real β-t crystal, which is much larger than this size, is probably obtained through nonequilibrium growth based on a small β-t nucleus. Since the surface energy of β-t boron is rather isotropic, it will probably not have a wire-like ...

The surface energies of α-rhombohedral, β-rhombohedral, and α-tetragonal boron were calculated from first principles, for the first time, to investigate their role in nanoscale crystal growth. The results support the existence of the pure α-tetragonal polymorph of boron. Although the existence of this polymorph had long been denied, α-tetragonal boron was recently prepared in the form of a nanowire. Pure α-tetragonal boron, despite its low cohesive energy, is more stable than the other structures as a result of its low surface energy when the number of atoms is less than about 254. Because the size of the nanowire that was obtained experimentally was larger than this, the nanowire may be in a metastable state. The difference between the surface energy of the ab-plane and that of the ac-plane explains why the nanowires grow in the c direction.

Adsorption of H2, 2H2, O2, and CO on ZrB2(0001) was investigated by using high-resolution electron energy loss spectroscopy. In hydrogen and oxygen adsorption, a strong loss peak appears in the specular spectra, showing the dissociative adsorption on the threefold hollow site. The vibrational frequency shifts toward higher energy with increase in oxygen coverage, while little changes in vibrational frequencies were observed in the hydrogen or deuterium adsorption. This means that the interaction between the adsorbates is stronger in oxygen adsorption than in hydrogen adsorption. The initial stage of CO adsorption shows two vibrational peaks, one of which corresponds to the atomic oxygen described above, the other is therefore the atomic carbon, showing dissociative adsorption.

The structure of the hafnium diboride (0001) surface was studied by impact-collision ion scattering spectroscopy. It was found that the clean surface, which is terminated by a flat (1×1) Hf layer at room temperature, turns into a surface with Hf adatoms after heating over 2000°C. This is probably caused by desorption of boron from the subsurface.

In order to investigate the mechanism of collision-induced electronic excitation, low-energy Ne + and Ne 0 have been scattered from an MgO(001) single-crystal surface together with clean and oxygenated polycrystalline Mg surfaces. One-electron excitation occurs efficiently during Ne 0 –Mg collision, which leads to (re)ionization of Ne 0 . At the MgO(001) surface, Ne 2+ ions are also created from Ne 0 incidence along the [110] azimuth where incident Ne 0 can make a sequential double collision on the aligned Mg 2+ ion rows, whereas Ne 2+ is not created in scattering along the [100] azimuth, where the Mg and O ions are arranged alternately. This result clearly shows that Ne 2+ is descended only from Ne + via one-electron excitation and not directly from Ne 0 via simultaneous two-electron excitation. From these facts, it is concluded that the unusually large Ne + yield from Mg should be ascribed to reionization of Ne 0 rather than autoionization of Ne**.

Vibrational frequency of H and D adsorbed on IVa and Va transition-metal carbide (TMC) (111) surfaces has been studied by high-resolution electron energy loss spectroscopy. Hydrogen (deuterium) was adsorbed dissociatively, showing the vibrational mode in an energy range of 110–120 (80–88) meV on ZrC(111), HfC(111) and NbC(111) surfaces, but at a much higher energy of 208 (149) meV on TaC(111). The former is similar to the case of H/TiC(111) consistent with three-fold hollow site adsorption. The latter seems too high to be explained by a variation of the force constant. It seems likely that on TaC(111), hydrogen is adsorbed on a different site from the other H/TMC(111) systems.

The atomic structure of magnesium oxide on the TiC(001) surface was investigated using time-of-flight impact-collision ion scattering spectroscopy. A hetero-epitaxial MgO layer was formed by thermal evaporation of magnesium onto the TiC(001) surface followed by exposure to oxygen at room temperature. From the time-of-flight spectrum which sees deeper layers from the top surface, it is found that the Mg and O atoms are located on the on-top site of the TiC(001) surface and the thickness of the MgO layer is restricted to two layers.

Tungsten diboride, WB2, usually has a hexagonal structure with the space group P63/mmc (number 194); and molybdenum diboride, MoB2, has a trigonal structure with R3̅m (number 166). Other than these phases, both diborides are reported to have a phase with an AlB2-type structure (P6/mmm, number 191). AlB2-type MoB2 is easy to synthesize and has been extensively studied, whereas AlB2-type WB2 is very difficult to synthesize and has appeared only once in a report by Woods et al. in 1966 (Woods, H. P.; Wawner, Jr., F. E.; Fox, B. G. Science1966, 151, 75.) We have investigated these diborides by means of first-principles calculations and found that boron defects are responsible for the difference in their synthesizability. AlB2-type MoB2 became stable enough with some boron defects added, while AlB2-type WB2 became minimally stable, suggesting it may not actually exist. Following our calculations, we attempted to synthesize AlB2-type WB2 with the optimum quantity of boron defects but observed no trace of it. We conclude, from both calculations and experiments, that AlB2-type WB2 does not exist stably in the W-B phase diagram and that the compound produced in Woods et al.'s report might have contained some impurities.

First-principles DFT calculations indicate that boron defects are responsible for the difference in synthesizability of the title compounds.

Low-energy (10 eV - 500 eV) He + ions are scattered from the LaB 6 (001) surface and the ejected electrons are detected. Two characteristic peaks are discernible in the electron energy spectra ; one is assignable to the surface state of the LaB 6 (001) surface resulting from Auger deexcitation of He * ( 3 S) descended from He + , and the other is due to autodetachment of temporary He - ions. These two peaks are very little broadened with increase of the velocity perpendicular to the surface, indicating that the electron emission occurs rather distant from the surface. On the other hand, marked energy broadening depending on the target species has been reported to occur if the electron is ejected via the Auger neutralization process. This is because Auger neutralization takes place even in a closer encounter with the individual surface atoms.

Low-energy (0.1–2 keV) helium and neon have been scattered from the Al(111) surface to investigate the mechanism of electronic excitation. Neutral He 0 and Ne 0 are (re)ionized during collision with Al as a result of one-electron excitation, whereas autoionization occurs preferentially for Ne + or Ne 0 incidence due to creation of two core vacancies. It is found that the Ne 2+ ion is formed during collision with Al and its intensity is largely dependent on the charge state of the primary beams; both single-scattering and double-scattering peaks of Ne 2+ emerge from Ne + incidence, whereas the double-scattering peaks of Ne 2+ dominates in Ne 0 incidence. This is because Ne 2+ originates from Ne + via one-electron excitation provided that the initial vacancy of Ne + exists in the 3dπ orbital and another vacancy is newly created in the 4fσ orbital after crossing the band. The predominance of one-electron excitation is characteristic of surface scattering and is basically caused by the band effect of energy-level crossing.

Room-temperature oxygen adsorption on NbC(111) and WC(1010) was studied by high-resolution electron energy loss spectroscopy and ultraviolet photoelectron spectroscopy. On NbC(111), oxygen is adsorbed dissociatively at the first stage. Above 2 X 10 -4 Pa s of exposure, a molecular species appears, which shows O-O stretching and M-O-O bending vibrations at 122 and 89 meV. These peaks disappear at 300°C. When the NbC(111) surface is saturated by dissociative N atoms prior to the oxygen adsorption, the molecular adsorption occurs from the first stage. This suggests that the molecular adsorption is caused by a site-blocking mechanism. On the other hand, oxygen is adsorbed molecularly on WC(1010) from the first. The UPS shows a broad peak at - 6.5 eV from the Fermi level, which is consistent with the 1π u and 3σ g orbitals of the molecular oxygen species.

The structures of oxygen-adsorbed and deuterium-adsorbed NbC(111) surfaces have been studied by impact-collision ion-scattering spectroscopy (ICISS) reflection high-energy electron diffraction (RHEED) and by computer simulation. At room temperature, both oxygen and deuterium are dissociatively adsorbed on the clean surface without long-range order, and an adsorbate atom goes into the threefold hollow site. The saturation coverage is more than 0.9 ML (monolayer) in both cases. The height of an adsorbate from the first layer is 1.10 A for oxygen and 1.00 A for deuterium, which corresponds to 2.13 A for the ONb distance and 2.08 A for the DNb distance. The ONb distance a little larger than the sum of the covalent radii (2.00 A) and the DNb distance is much larger than the sum of the covalent radii (1.64 A). Judging from the bond length and electronegativity, O and D atoms are supposed to have some negative charge.

Low-energy H + ions have been scattered from clean and oxygenated TiC(001) surfaces to investigate the mechanism of dynamical electronic transitions. The H + spectrum exhibits a marked surface peak of Ti, which is ascribed to reionization of neutralized H + rather than survival of the incident H + ions. Upon oxygenation, the H + (H − ) yield is increased (decreased) considerably; the surface peak for oxygen is almost absent in the H − spectra although it is remarkable in the H + spectra. These results are qualitatively elucidated on the basis of the model in which the electron transitions are mediated by the molecular orbital during a violent collision or a short-lived chemisorption state of hydrogen on the surface. The H 1s orbital has a strong antibonding character when colliding with the negatively charged oxygen atoms, by which the probabilities of neutralization, reionization, and negative ionization of the proton is largely affected after oxygenation.

In order to investigate the mechanism of electronic excitation during rare-gas ion scattering from surfaces, molecular orbital energies of various diatomic collision systems have been calculated on the basis of the discrete variational X α method. Doubly-excited autoionization is caused by the highly promoted 4fσ orbital of Ne colliding with the third-row elements of the Periodic System (NaAr), with decrease of the electron transition probability for heavier target atoms. By contrast, one-electron excitation leading to reionization is basically caused by the pseudo-crossing of the molecular orbitals. This is the reason why reionization is observed experimentally for a larger variety of projectile-target combinations compared to autoionization. Reionization may not be predicted sufficiently by the conventional treatment based on the diabatic correlation diagram. It is found that the probability of reionization is suppressed if the ionic level has a bonding character due to the interaction with the target d orbitals.

The formation of monolayer graphite on a tungsten-segregated TiC(1 0 0) surface is reported. This is of particular interest since monolayer graphite can not be easily formed on the pure TiC(1 0 0) surface. The monolayer graphite on the tungsten-segregated surface is identified by low-energy electron diffraction and its phonon dispersion relations are determined by high-resolution electron energy loss spectroscopy.

Neutralization and electronic excitation during low-energy ion-surface scattering have been investigated from a combination of experiments and molecular-orbital energy calculations based on the discrete variational X\ensuremath{\alpha} method. It is found that a variety of electronic transitions are mediated by the molecular orbitals during the violent collision or the short-lived chemisorption state of the ions, which may not be inferred from the atomic orbitals of the isolated projectiles. The rare-gas ions, such as ${\mathrm{He}}^{+}$ and ${\mathrm{Ne}}^{+}$, capture a valence electron mainly via the Auger neutralization process, whereas resonant tunneling can also play an important role in neutralization of the reactive ions (${\mathrm{H}}^{+}$, ${\mathrm{N}}^{+}$, ${\mathrm{O}}^{+}$). The occurrence of resonant tunneling is related to the open-shell structure of the reactive ions. The probability for resonance neutralization is sensitively dependent upon the nature of the valence band (band effect) and is largely enhanced at the metal surface relative to the ionic-compound surface. The valence-band electron can be excited in scattering of reactive ions as a sequence of neutralization/ionization along the promoted antibonding molecular orbital. In scattering of the rare-gas ions, not only the valence electron but also the semicore electron can be excited for specific ion-target combinations in which sufficient promotion of the antibonding molecular orbital can take place during collision.

The mechanism of neutralization of and negative ionization from hydrogen ions is studied via experiments in which H + beams are incident on LiCl surfaces, combined with molecular-orbital-energy calculations. It is found that charge transfer and electron excitation occur due to the resonant-tunneling and electron-promotion mechanisms via a short-lived chemisorption state of hydrogen. The probability for H + to H − conversion at the LiCl surface depends on the species of the target atoms and is enhanced in collision with Li + by one order of magnitude relative to Cl − .

Low-energy ${\mathrm{D}}^{+}$ scattering is employed to explore the nature of the bonding of polycrystalline alkaline-earth oxides MgO, CaO, SrO, and BaO, with particular emphasis on the investigation of the ionicity of the topmost-layer atoms. Increasing ionicity as one goes to the heavier cations is concluded from the probability of the resonance neutraliztion of the ${\mathrm{D}}^{+}$ ions, which is consistent with the conventional chemical arguments based on electronegativity scales but is in apparent contradiction to the results of recent ab initio cluster-model calculations. It is also concluded that the metallic Ba layer is formed rather patchily on the BaO surface after the heat treatment up to 1000 \ifmmode^\circ\else\textdegree\fi{}C. This is probably because free Ba atoms, being supplied by the reaction of BaO with the Ta substrate, are precipitated at the BaO surface. Another example is concerned with the interactions of the Ba adatoms with Si(001) and Pt(111) surfaces; Ba is found to have marked covalency with the substrate atoms.

The segregation of W impurities at the TiC(001) surface has been investigated by using impact collision ion scattering spectroscopy and Auger electron spectroscopy. The topmost layer of TiC(001) is strongly enriched with W accompanied by carbon vacancies (20–25%) in their nearest neighbors, while no marked enrichment of W occurs in the deeper layers. The oxygen is incorporated by the (001) surface irrespective of the existence of W segregants, whereas its uptake rate is moderately enhanced by W, due probably to oxygen diffusion into the bulk mediated by the carbon vacancies.

We report on the systematic investigation of neutralization of low-energy D[sup +] ions scattered from 17 ionic-compound surfaces with particular emphasis on the relation between the neutralization probability of D[sup +] and the ionicity of the surface atomic bonds. The D[sup +] ions tend to survive neutralization provided that the target atom is completely ionized, whereas almost complete neutralization results if the target has covalent or metallic orbital hybridization with the ligand. The D[sup +] spectra from nitrates, sulfates, and carbonates of Na and K, as well as K[sub 2]SiO[sub 3], exhibit surface peaks of D[sup +] surviving neutralization for both cations and oxygen, indicating large ionicity in the alkali-metal--oxygen bond. Rather exceptional is Na[sub 2]SiO[sub 3], for which a relatively small ionicity is concluded from the absence of surface peaks. For the alkaline-earth compounds, ionicity decreases in going to the light elements as evidenced by the decrease of the surface-peak intensity. The occurrence of charge-exchange relaxation between cations and anions of NiSO[sub 4] and CuSO[sub 4] results in large neutralization probability of D[sup +], while the closed-shell structures of Ag[sup +] (4[ital d][sup 10]) and Pb[sup 2+] (6[ital s][sup 2]) in AgNO[sub 3] and Pb(NO[sub 3])[sub 2], respectively, suppressmore » neutralization to some extent so that the appreciable surface peaks appear.« less