Search

Ultrasonic surface rolling process (USRP) is one of the effective mechanical surface enhancement techniques. During the USRP, unstable static force will easily do harm to the surface quality. In order to achieve a higher surface quality on the part with a curved surface, an active and passive compliant USRP system has been developed. The compliant USRP tool can produce the natural obedience deformation along the part surface. Force control based on the fuzzy Proportional-integral-derivative (PID) method is then designed to maintain the static force during the USRP. Experiments have been performed on a real aero-engine blade with curved surface. It is proved that the deigned active and passive compliant USRP system can significantly reduce the force variation from 42.2 N to 4.2 N, and achieve a uniform surface quality after processing.

Double-sided tools system is a promising design to counteract the forces caused by tools, and has become a practical way to machine thin-walled parts with the twisted structure. However, it is still difficult to obtain uniform machining effects on the part surfaces. This paper proposes a coordinated motion planning method for coordinating two tools in the double-sided tools system. The tools are devoted to guaranteeing the uniform surface quality while meeting the constraints about tools' velocity, tools' inclination and the system structure. At first, a velocity-dependent model is presented to describe the two tools' motions along the predefined paths. Then, the motion-planning problem of two tools is formulated as a constrained velocity-optimization problem, and the genetic algorithm is proposed to achieve the optimization target of the surface uniformity. Finally, experiments are performed on a real blade in a self-developed double-sided ultrasonic surface rolling process machine tool. The results demonstrate that the proposed coordinated motion planning method can effectively improve the surface uniformity in the double-sided tools system.

As crucial parts of an aeroengine, blades are vulnerable to damage from long-term operation in harsh environments. The ultrasonic surface rolling process (USRP) is a novel surface treatment technique that can highly improve the mechanical behavior of blades. During secondary machining, the nominal blade model cannot be used for secondary machining path generation due to the deviation between the actual and nominal blades. The clamping error of the blade also affects the precision of secondary machining. This study presents a two-sided USRP (TS-USRP) machining for aeroengine blades on the basis of on-machine noncontact measurement. First, a TS-USRP machining system for blade is developed. Second, a 3D scanning system is used to obtain the point cloud of the blade, and a series of point cloud processing steps is performed. A local point cloud automatic extraction algorithm is introduced to extract the point cloud of the strengthened region of the blade. Then, the tool path is designed on the basis of the extracted point cloud. Finally, an experiment is conducted on an actual blade, with results showing that the proposed method is effective and efficient.

Security and performance of aircraft engines are extremely affected by the blades’ surface quality. Ultrasonic surface rolling process (USRP) can improve material surface properties effectively and increase the alloys’ service life. However, applying USRP onto blades remains a challenge due to the free-form surface with thin-walled leading/trailing edges and the pre-load pressure on the blade during the USRP process. In this paper, a coordinated bilateral method of USRP on the blade is proposed. A machine tool is first designed and integrated with two USRP devices. A method of coordinated bilateral path generation is then proposed, so that the blade surface can be processed by the two USRP devices simultaneously. Finally, we perform experiments on a real aero-engine blade. The results demonstrate that by using the proposed method, USRP can be applied on the bilateral surfaces of the aero-engine blade coordinately.

Assembly-line polyketide synthases, such as the 6-deoxyerythronolide B synthase (DEBS), are large enzyme factories prized for their ability to produce specific and complex polyketide products. By channeling protein-tethered substrates across multiple active sites in a defined linear sequence, these enzymes facilitate programmed small-molecule syntheses that could theoretically be harnessed to access countless polyketide product structures. Using cryogenic electron microscopy to study DEBS module 1, we present a structural model describing this substrate-channeling phenomenon. Our 3.2- to 4.3-angstrom-resolution structures of the intact module reveal key domain-domain interfaces and highlight an unexpected module asymmetry. We also present the structure of a product-bound module that shines light on a recently described “turnstile” mechanism for transient gating of active sites along the assembly line.

The aim of this paper was to investigate the effect of thermal annealing on the microstructure, mechanical properties, and residual stress relaxation of deep rolled pure titanium. The microstructure and mechanical properties of the surface modified layer were analyzed by metallographic microscopy, transmission electron microscope and in-situ tensile testing. The results showed that the annealed near-surface layer with fine recrystallized grains had increased ductility but decreased strength after annealing below the recrystallization temperature, where the tensile strength was still higher than that of the substrate. After annealing at the recrystallization temperature, the recrystallized near-surface layer had smaller grain size, similar tensile strength, and higher proportional limit, comparable to those of the substrate. Moreover, the residual stress relaxation showed evidently different mechanisms at three different temperature regions: low temperature (T ≤ 0.2 Tm), medium temperature (T ≈ (0.2‒0.3) Tm), and high temperature (T ≥ 0.3 Tm). Furthermore, a prediction model was proposed in terms of modification of Zener-Wert-Avrami model, which showed promise in characterizing the residual stress relaxation in commercial pure Ti during deep rolling at elevated temperature.

Depth video contains depth and disparity information of a scene, which is critical for 3D video systems. In this paper, a convolutional neural network (CNN) based block upsampling method is proposed to improve the efficiency of depth video intra coding. For each largest coding tree in a depth map, it is down-sampled before sent into encoder and recovered into the original size in an intelligent way after low-resolution coding. A novel texture-assisted CNN (TACNN) is presented to handle the depth block up-sampling. The network is made up of several residual coding units and the features of texture block are extracted to assist the reconstruction of the corresponding depth block. Experimental results show that the proposed method achieves competitive rate-distortion performance compared with the state-of-the-art approaches.

The surface quality of blades affects the performance and security of aircraft engines. Ultrasonic surface rolling process (USRP) can be used to improve service life of alloys. However, blades with complex design specifications are difficult to processing, especially the thin-walled leading/trailing edge should bear the loads during the process of ultrasonic rolling. Therefore, it's still a challenge to apply USRP for the surface strengthening of blades. In this paper, an integrated system that combines USRP and a multi-axis machine tool is presented to improve the surface quality of blades. Further, the control of the machine tool is reasonably designed. Finally, the proposed method is validated by an experiment on a real blade. This proves that the system can be effectively used for automated strengthening of aircraft blades.

With the development of industry and commerce, the demands for indoor positioning are increasing. Indoor positioning requires higher reliability, speed and accuracy of the positioning system. However, a single positioning system has its own advantages and disadvantages, which makes it difficult to meet the positioning requirements at the same time. Therefore, this paper proposes a fusion positioning technology that combines ultra-wideband UWB (Ultra-wide Bandwidth) positioning technology and inertial navigation technology using Kalman filter in the case of line of sight. Through the simulation experiment on matlab, by comparing fusion positioning technology with the single UWB positioning technology and the single inertial navigation technology, it is found that the fusion positioning technology improves the positioning accuracy, and meanwhile increases the stability of the system.

In this paper, under two different electromagnetic modes, the photonic band gaps (PBGs) in the two-dimensional plasma photonic crystals (PPCs) are theoretically investigated based on the plane wave expansion method. The proposed PPCs are arranged in rhombus lattices, in which the homogeneous unmagnetized plasma rods are immersed in the isotropic dielectric background. The computed results showed that PBGs can be easily tuned by the angle of rhombus lattices, and a cutoff frequency and a flatbands region can be observed under the TM and TE polarized waves, respectively. The relationships between the relative bandwidths of first PBGs and the parameters of PPCs in two such cases also are discussed. The numerical simulations showed that the PBGs can be manipulated obviously by the parameters as mentioned above. The proposed results can be used to design the waveguide and filter based on the PPCs.

Blades are one of the key parts in aircraft or turbine engine. Machining, such as polishing and grinding are frequently used to enhance the physics performance of blades to improve the service life. However, due to the free-form surface, it is still a challenge for path planning of machining on the blades. This paper proposes a path planning method of machining on surface of the blade. The machining path is first planned on the basis of contour curve of blades. The machining areas are then recognized according to the radius of curvature along the curves. Finally, a six-axis machine tool is developed, and experiments of surface machining are performed on a real blade. The results prove that the proposed path planning method for machining on the blades is effective.1

The optical millimeter-wave has been theoretically and experimentally investigated based on frequency-eightfold by only utilizing a single-arm Mach-Zehnder modulator and simultaneous optical carrier suppression. The suppression of central carrier and residual orders is realized via adjusting the parameters of Mach-Zehnder modulator, including the drive signal that amplitude voltage by the radio frequency signal and the bias voltage of direct current. We experimentally generate the 72-GHz optical millimeter-wave, which adopted frequency multiplication by a radio frequency signal of 9-GHz, is not using any costly filters. Furthermore, we realized the 3.5-Gb/s OOK signal for downlink of radio-over-fiber system. In which by the generation of 72-GHz millimeter wave. To the best of our knowledge, the paper realized generation of frequency eightfold millimeter wave signal and suppressed the undesired carrier at the same time via just utilizing one single arm Mach-Zehnder modulator in our scheme for the first time.

The accuracy of demand forecasting is of great importance for automobile manufacturing, transportation and selling. Considering automobile demand much more sensitive to price waving and customer’s purchasing decision deeply relied on internet searching information, this paper offered an integrated model of PLS regression and ANFIS for the automobile demand forecasting. To simplify the problem, we use partial least square (PLS) regression to select the most influential factors first including sales price, advertising cost, searching index, consumer satisfaction index. Then these factors and sales data are taken as the input variables into an adaptive-network-based fuzzy inference system (ANFIS) to get the forecast. To test and validate the proposed method, we compare our model with Autoregressive Integrated Moving Average (ARIMA) model and Time Series Decomposition Multiplicative (TSDM) model using sales data collected from Changchun. The results show that the proposed model outperforms the other two models, especially when the market environment fluctuates dramatically.

The lattice dynamics of wurtzite GaN and AlN is studied within the framework of a rigidion model. Short-range interactions up to the third nearest neighbors are described by using a valence-force-field potential and the long-range Coulomb interactions between ions are calculated via an Ewald summation. Phonon dispersion curves, density of states and specific heats are presented.

Band structures of Si-C and Si-C-Ge alloys are calculated by using the ab initio LMTO (linear muffin-tin orbital) method within the framework of atomic-sphere approximation. The effects of different atomic configuration and lattice relaxation on the band structure of alloys are taken into account. The results show that for large concentrations of C in Si and Si-Ge, the band gap increases monotonically, while for small concentrations of C in Si and Si-Ge, the band gap shrinks. The lattice relaxation further reduces the band gap. The possible explanations for the reduction of band gap of Si-C and Si-C-Ge alloys are presented.

Trajectory simulations are carried out for vibrationally assisted adsorption of H2 on Cu(100) surface. The potential-energy surfaces for H2 approaching the copper surface in any molecular orientation are obtained by using the embedded-atom method. It is clearly shown that the vibrational energy of incident H2 greatly promotes the dissociation by assisting H2 to surmount the activation barrier at the reaction zone. The dependence of initial sticking coefficient S0 for H2 activated adsorption on Cu(100) surface on the normal incident translational energy EN has been calculated. The different contributions from H2(ν = 0) and H2(ν = 1) to the sticking probability of H2 molecular beam adsorption on Cu(100) are investigated. The results are in well agreement with the experimental observations and other theoretical calculations.

A theoretical model for describing H2 dissociative chemisorption on Cu surfaces is proposed. The sticking probability S is calculated as a function of vibrational state, average kinetic energy and incident angle of hydrogen molecular beam. Within the theoretical frame of this model, the different contributions to S from H2(ν = 0) and H2(ν = 1) can be clearly distinguished. The calculated results indicate that vibrational energy significantly promotes the chemisorption of H2 on Cu surfaces in the region of low translational energy. The equations derived can be used to analyze the experimental data for both pure and seeded molecular beams.

The electronic and structural properties of the Si/GaP(110) interface with various thicknesses of Si overlayers on a GaP(110) semi-infinite substrate have been studied by using the self-consistent semi- empirical tight-binding method with the charge neutrality condition. In addition, we present an approach to calculating the variation of band offsets with the number of adlayers during the Si/GaP(110) interface formation. In the calculation, the influence of surface and interface dipoles are taken into account. The calculation shows that the band offset exists with the first Si adlayer on the GaP(110) substrate and increases continuously with the number of Si overlayers. When reaching five or six Si layers, the band offset is close to the final value, 0.43 eV, of the Si/GaP(110) interface of two semi-infinite crystals.

The dissociative adsorption of H2 On Cu(100) is studied by using the embedded-atom method (EAM). The molecular orientation dependence and the impurity effects have been investigated. It is found that the most favourable molecular orientation for H2 dissociation is to keep the H-H axis parallel to the surface. In this case, the corresponding activation barrier is lowest. Calculations show that when an impurity atom is present in the Cu substrate, the chemisorption properties near the impurity atom are affected. Ni and Pd can promote the H2 dissociation by lowering the activation barrier height; Ag, however, impedes the H2 dissociation since the activation barrier is lifted. Various H2 dissociation pathways are discussed and the corresponding potential-energy surfaces (PESs) are presented.

The dissociative adsorption of a hydrogen molecule on the nickel(100) surface with point defects is investigated using the embedded-atom method (EAM). The potential-energy surfaces (PES) for H2 dissociation on both perfect and imperfect Ni(100) surfaces are presented, based on total-energy calculations. it is clearly shown that as the H2 approaches the Ni(100) surface along the entrance channel, the H-H bond is progressively weakened while the H-metal bonds begin to form; finally the H2 is adsorbed on the surface in the form of two independent H atoms. This dissociation process is affected by the vacancy and impurity atoms existing in the Ni substrate. The activation barriers (Ea) for the dissociation of H2 through various pathways are calculated. The barriers for the dissociation of H2 on the perfect Ni(100) surface are found to be low (about 0.08-0.09 eV. corresponding to different dissociation pathways). The existence of vacancies enhances the dissociation of H2 by lowering the activation barrier height and providing more adsorption sites. However, the impurity atoms (Cu, Pd) can impede the dissociation of H2 on the Ni(100) surface by increasing the activation barrier height. The adsorption heat of H2 chemisorption on the contaminated Ni(100) surface is also calculated. It is found that the effects of impurities on the dissociation of H2 vary with the dissociation pathways and the impurity sites.

The Carlsson-Gelatt-Ehrenreich inversion technique is applied to the derivation of pair potentials for the elemental bonds C-C and Si-Si, and the compound bond Si-C. Chen's Mobius function method is used to speed up the convergency. The cohesive energies are obtained from the ab initio pseudopotential total energy calculation of Chang and Cohen within the framework of the local-density approximation and fitted to the universal binding energy function of Rose et al. The pair potentials thus obtained reproduce exactly the full cohesive energy curves, lattice constants and bulk moduli.

We give the correct dynamic matrix elements for the diamond structure up to the sixth neighbors in Born-von Karman model. The importance of the fifth neighbor to the characteristic flattening of the transverse-acoustic (TA) modes in Ge and Si is confirmed by extensive Monte Carlo iterative search of the parameter space. This indicates that range up to at least fifth neighbor should be considered when constructing potential models for Ge and Si in order to properly produce the lattice dynamics.

Structural behaviors of silicon and germanium clusters under the recently proposed modified Stillinger-Weber (SW) potential have been studied by molecular dynamics method along with the conjugate-gradient optimization. Since the SW potentials have been derived from the properties of bulk materials, it gives relatively large discrepancy when the cluster number is small, e. g. n=3 and 4. When n>4, the ground-state structures under the modified SW potential are close to that from the first-principle calculation. The binding energies are also improved under the modified SW potential. These results may be attributed to the relative enhancement of the two-body term over the three-body term in the modified SW potential, which leads to structures with a preferential coordination number 4. Structural behaviors of germanium clusters are similar to those of silicon.

The linear muffin-tin orbital method in the tight-binding representation is adopted to study the electronic structures of the native defects in beta -SiC with the Green function approach. The original potential parameters are obtained by the standard band calculation of bulk beta -SiC and the change in the potential parameters is obtained self-consistently. It is found that the silicon vacancy acts as an acceptor and the carbon vacancy acts as a donor with the split-off states at 0.45 and 1.67 eV above the valence band maximum. For the antisite defects, there is no split-off state in the band gap.

The linear muffin-tin orbital method is used with the atomic-sphere approximation to study the electronic properties of beta -SiC. The equilibrium lattice constant, bulk modulus and band structures calculated are in good agreement with the experimental results. The strain effect on the band structures is discussed; the results show that, when the lattice constant parallel to the surface is expanded by 8% or more the indirect fundamental gap might become direct.

The linear muffin-tin orbital method is used with the atomic-sphere approximation to study the ground-state properties, band structures and pressure dependence of the band gaps of zincblende GaN. The equilibrium lattice constant calculated is smaller than the experimental value by about 2%. It is found that the fundamental band gap is direct from Gamma 15v to Gamma 1c and that the value (3.90 eV) is in good agreement with the experimental value. The linear pressure coefficient for this band gap is much smaller than those for GaAs and GaP. The total density of states and local density of states are discussed. Moreover, the band structure of wurtzite GaN is also studied. It is found that the band gap of wurtzite GaN is smaller than that of zincblende GaN, which agrees well with the experimental results.

The adsorption properties of Au on β-SiC(111) surfaces are studied using the charge self-consistent extended Huckel theory. The intermixing between Au atom and β-SiC could happen only on the Si-terminated (111) surface. In the case of Si-terminated, there is a intrinsic surface state in the energy gap, and the Fermi energy is pinned at this surface state. But in the case of C-terminated, the surface state exists near the top of the valence band.

The Keating model is employed to determine the geometric structure of the (2*1) reconstructed beta -SiC(100) surface. The symmetric dimerization for both the Si- and C-terminated (100) surfaces is obtained by minimizing the elastic energy. The extended Huckel band calculation is performed for the electronic structures of both the ideal and the (2*1) reconstructed beta -SiC(100) surfaces. The results show that the ideal surfaces are metallic, whereas the (2*1) reconstructed surfaces are semiconducting. Moreover, the band structures calculated agree quite well with the experimental results. The density of states is also calculated and discussed.

The validity of the Stillinger-Weber (SW) potential models for silicon and germanium is tested by using lattice-dynamics calculations. It is found that the phonon-dispersion relations described by the SW potentials are not satisfactory. A new set of parameters is proposed that provide considerable improvements for the phonon-dispersion relations of both Si and Ge. In addition, some applications of the modified SW potentials are given.

Determination des geometries de ce reseau contraint, pour differents substrats. Pour un choix correct de la composition x, le superreseau (Si) 4 /(Ge) 4 symetriquement contraint peut etre obtenu. Les contraintes sont prises en compte dans les etudes de la dynamique reticulaire du superreseau

Lattice dynamics of strained (Si)4/(Ge)4 superlattice grown pseudomorphically on (001)-oriented Si1-xGex (0 ≤ x ≤ 1) substrate is investigated. In the present calculations, the effects of strain and substrate are discussed.

A dynamical calculation of the state equation is presented to give a good prediction of the melting temperature for silicon which is much better than the molecular dynamics simulation result and agrees satisfactorily with the experimental value.

Using first principles total energy calculations,the surface energies of Mo and W before and af-ter adsorbing Rh thin films in [111] and [110] orientations are calculated.The calculation results indicate that clean (111) surfaces of both Mo and W cannot transform to {110} facets,in agreement with experimental observations.When the coverage of Rh arrives at one physical monolayer,Rh/Mo(111) still cannot facet to {110},while under a favorable energy condition for the transforma-tion,Rh/W(111) may facet to {110}.

The vibrational properties of spherical nanocrystal Si were studied and the relation between the size and Raman shift was established. It was found that the size of porous Si, simulated by Si spheres, should be much smaller than commonly reported.

The electronic energy bands of a series of β phases of carbon nitride, silicon nitride and germanium nitride have been calculated using a first-principle linear combination of muffin-tin orbital method with the atomic-sphere approximation(LMTO-ASA). The band gaps obtained for the three compounds are 4. 1751,5. 1788 and 4. 0279 eV, respectively. For β-C3N4 ,the nitrogen 2p orbitals plays an important role in the top of valence band(VB) due to its non-bonding. For -β-Si3N4,the silicon 3d orbitals plays an important role both in the VB and in the conduction band(CB). Its band gap is indirect which is in agreement with other calculated results. For β-Ge3N4,the germanium 3d orbitals preserve the characteristics of atomic orbitals and give very little contributions to other VBs and CB. Its 4d empty orbitals play a role analogue to 3d of Si of β-Si3N4. This is the frist theoretical study reported on the electronic structure of β-Ge3N4.

The electronic structures of short period (Si)n/(Ge)m Strainedlayer superlattices (SLS's) are calculated using empirical tight-binding method. The results indicate that due to the zone-folding, only the SLS's with n+m=10 have the posibility to form direct band gap. For SLS's with n+m=10, the type of energy gap is determined by the relative energies of conduction band valleys at Γ, N, and Δ. However, the Ge layers number and the composition of substrate play an important role. For (Si)6/(Ge)4 and (Si)8/(Ge)2 SLS's grown on Si1-xGex substrate, direct or quasidirect band gap could be got for the substrate composition 0.4≤x≤1.0 For (Si)4/(Ge)6, direct gap could be obtained only for 0.3≤x≤0.6, and in (Si)2/(Ge)8 SLS's the gap could not be direct for all substrate composition.

Deep levels of vacancy and substitutional impurity atoms in strained Si or Ge grown on alloy substrates are investigated. The band structures of strained bulk are calculated by using the empirical tight-binding method. The method of Green's function has been used to calculate the defect levels. The results show that the triplydegenerate p-like T2 level of defect in bulk Si or Ge splits into two levels due to the strain. The value of splitting increases with strain increasing. Owing to the strain,the valence band maximum of Si or Ge shifts upward, hence some deep levels in the bulk might transform to resonant ones.

The understanding of chemical reactions which occur at the interface between metal and semiconductor is of both practical significance and theoretical interest. However, due to the difficulties of studying solid-solid reaction on an atomic scale, the mechanism responsible for those reactions remains somewhat unclear. There still exist some controversies regarding whether metal-semiconductor junctions are abrupt or not. In the present article, chemisorption geometries and electronic structures of group III metals, noble metals, transition metals and rare earth metals on semiconductor surfaces are reviewed and discussed from the theoretical point of view. Since most extensive experimental studies have done on Si(111) and GaAs(110) surfaces, more detailed discussions will be devoted to chemisorptions on those cleaved surfaces, such as Al/Si(111), Al/GaAs(110), Ag/Si(111), Ni/Si(111), Mo/Si(111) and interfaces of transition metal silicide and silicon. For studying chemisorbed surfaces and interfaces both slab models and cluster models are used, and results from both the first principle method and the empirical method are presented. In many cases, where local properties are emphasized, a cluster consisting of a finite number of atoms is quite adequate for simulating a semi-infinite surface. Although the first principle calculation is more accurate, however, it is quite expensive and time consuming, and the empirical method still can offer reasonable results for comparing with data from experiments.

The monolayer adsorbed Mo(001) surface is studied by the slab dynamical calculation by using the nearest and next nearest neighbour interactions. The phase diagram is obtained for light, heavy atom adsorbed and clean Mo(001) surface. The surface reconstruction induced by the interactions between adsorbed atoms is discussed.

A theoretical investigation of NiSi2/Si(111) interfaces using empirical tight binding method and Slab Model is presented. The positions of Fermi level Ef for various types of interfaces has been compared. The heights of Ef for type-I(a) and type-II(a) or for type-I(b) and type-II(b) are the same; but there exists a difference of 0.1eV between type-I(a) and -I(b) as well as for type-II(a) and -II(b). The local densities of states of some typical layers in the slab have also been studied.

Cluster model is used to simulate several optical vibration modes of oxygen in the layered perovskite La-Ba-Cu-O. It is found that only the breathing mode of oxygen induces a charge fluctuation between Cusites. The charge fluctuation is reasonably large and may have important implications in the high-Tc superconductivity.

Lattice dynamics of 36Ar are investigated by Keating model. Owing to the short-range character of the interactions in rare-gas solids, only the nearest neighbour interactions were taken into account. The calculated phonon frequencies and elastic constants are in good agreement with the experimental values. The maximum error for phonon frenquencies is 3.3%, and for the elastic constants is 7%.

The Si(001) surface is studied by the slab dynamical calculation using Keating model with the inclusion of an anisotropic pair interaction between surface atoms. T=0 phase diagram is obtained by diagonalizing the slab matrix numerically. Detailed information of the Si(001) surface reconstructive structure is afforded and to be compared with experiments.

The electronic states of MoSi2 in the body-centred tetragonal CaC2 structure have been studied using the self-consistent linear muffin-tin orbital method. The core states are treated in a fully relativistic way, whereas for the valence states a semirelativistic approximation is utilised. The results show that the majority of Mo d electrons interact with Si p electrons; therefore there exists a strong p-d hybridisation state in MoSi2. The energy band structure is compared with that calculated by the fully relativistic LCGO (linear combination of Gaussian orbitals) method; it was found that the spin-orbit coupling has little effect on the band structure of MoSi2. The total and the projected densities of states are also calculated and compared with the photo-emission data.

In the present work, the chemisorption of Ge on GaAs(110) was studied by a cluster model using the charge self-consistent extended Huckel method (EHT) and the results thus obtained were compared with those derived from the experiments.It was shown from the calculation that after chemisorption, the substrate tends to relax back to the ideal unrelaxed state, after cleavage. The tilted angle of the As-Ga bond was reduced from the well known 27° to 10°. The gap states which were driven into both the valence band and the conduction band due to relaxation will-be introduced into the gap after chemisorption. From the calculation of the total energy of the cluster, it was found that the Ge atom could bind to both Ga and As atoms although the binding of the latter was somewhat stronger. The gap states were derived from the p states of Ge and the s and p states of Ga and As atoms. The states will serve as donors and acceptors for p and n type materials respectively. From a comparison with the previous results concerning studies of group III and V metals on GaAs(110) surface, it seems that the change of surface relaxation after chemisorption might be the origin of gap states and the cause for Fermi level pinning. The explanation was first put forward by Chen et al. from their experimental studies.The present calculation also gives the core shifts of Ga(3d) and As(3d) of the clean relaxation surfaces with respect to that of the bulk as well as the core shifts of Ga(3d) and As(3d) after the chemisorption of Ge atom.A chemisorption induced state which is attributed to a precursor state for a heterojunction interface state is also obtained at a binding energy near 7eV which is in fairly good agreement with the experimental result.

For solving the long existing controversy about the chemisorption geometry of Cl/Ag(100), we show, by a SW-Xα-SCF cluster calculation, that the SOM (simple overlayer model) is more plausible. This conclusion agrees quite well with the more recent high resolution results obtained by angle-resolved photoemission.

The authors have performed LMTO ASA calculations of the electronic structures of LaM3 (M identical to Sn,In). It is shown that the interactions between La d and Sn or In p states are the main bonding states in LaM3, therefore indicating that LaM3 behaves like a transition-metal compound. The differences between the electronic structures of LaM3 and CeM3 are discussed. The lattice constants and bulk modulus of LaM3 are also calculated and compared with experimental values.

In the present paper both the Ge chemisorption on Si(111) and Si chemisorption on Ge(111) are studied by the cluster model and charge self-consistent extended Huckel method. The results show that the threefold hollow site is energetically favorable for both systems, while the electronic states are somewhat different with each other. The situation for Ge/Si(111) calculated agrees fairly well with that for the 1 3 ML deposition of Ge at room temperature observed by experiments.