Wrong Bonds Research Articles

Effect of nanophase separation on crystallization process in Ge–Se glasses studied by Raman scattering

We have found evidence of nanophase separation of the two structural units, Ge 2(Se 1/2) 6 ethane-like units and Ge(Se 1/2) 4-tetrahedra ones, in the Ge content range of 0.30–0.37 in Ge x Se 1− x glasses by investigating Raman spectra in a temperature range of 300–1000 K. The influence of nanophase separation over the crystallization processes induced by heating the bulk glasses beyond the glass-transition temperature, T g ’s , is explained in a medium-range-order structural point of view. A randomly distributed Ge 2(Se 1/2) 6 ethane-like unit induces an additional stress to the backbone. The local stressed structure relaxes through breaking the Ge–Ge bond (constituting the ethane-like unit) and acquiring a nearby Se atom (belonging to the Se–Se wrong bonds) to construct the corner-shared Ge(Se 1/2) 4-tetrahedra units. An increased nanophase separation promotes the nucleation of GeSe crystal in the super-cooled liquid state of Ge x Se 1− x ( x=0.36 and 0.37) in which we obtained the crystalline GeSe 2 and GeSe simultaneously, since a considerable population of ethane-like units lets them cluster in the glasses.

Annealing Twinning in Boron-Doped Ni<SUB>3</SUB>Al

Dependence of frequency of annealing twinning events on annealing temperature and composition, especially boron content, in polycrystalline Ni 3 Al alloys has been investigated by means of optical and transmission electron microscopy, and also by evaluation of coherent twin boundary energy via counting the wrong atomic bonds across twin boundaries. Microstructural observation found that after homogenization at temperatures from 1000 to 1360°C both binary and boron-doped Ni 3 Al alloys without prestrain show annealing twins only on the Al-rich side with respect to stoichiometry below 1200°C. Above 1200°C annealing twinning may also take place on the Ni-rich side of stoichiometry in the binary alloy system but no twins were found in the y + y' two-phase field. In contrast to plenty of annealing twins in binary alloys, addition of boron decreases greatly the number of twins. 0.2 at% addition significantly decreases the frequency of twinning events and no twins appear above 1200°C. 0.5 at% boron almost cleans out all the twins from the microstructure. Annealing twinning is supposed to be completed by two processes including sweeping of 1/3(211) superpartial dislocations on successive (111) planes and simultaneous diffusion of boron atoms into the energetically favorable octahedral interstice that are surrounded by six Ni atoms. Boron acts as a strong obstacle to the movement of the superpartial dislocations and greatly increases difficulty in annealing twinning.

Theoretical analysis of tilt grain boundaries in GaN at the atomic scale

ABSTRACTEpitaxial layers of GaN contain a very high density of threading dislocations. In the first stage of growth they may form low and high angle grain boundaries. Energetic calculations of <0001> tilt grain boundaries have been performed with the Stillinger-Weber potential modified to take into account the wrong bonds Ga-Ga and N-N. The variation of the energy has been calculated as a function of the rotation angle. Two minima exist with special atomic structures based on a limited number of structural units. They are used to describe the other misorientations in terms of dislocation cores.

Evidence of nanophase separation in Ge–Se glasses

Nanophase separation of two structural units, ethane-like Ge 2(Se 1/2) 6 and Ge(Se 1/2) 4 tetrahedra, in Ge x Se 1− x ( x=0.30–0.37) glasses has been investigated by Raman scattering measurements in a temperature range 300–1000 K. For x<0.34, randomly distributed Ge 2(Se 1/2) 6 ethane-like units induce an additional stress to the backbone of Ge(Se 1/2) 4 tetrahedral units. For x>0.34, a self-organization mechanism occurs in large regions of the nanoscaled-separated parts owing to: (i) a considerable concentration of ethane-like units, (ii) stresses within the regions, and (iii) lower constraints on the backbone network. For samples with x=0.36 and 0.37, each structural unit is stable below the glass-transition temperature, and crystalline regions of GeSe 2 and GeSe appear simultaneously. In the lower Ge-concentration sample, with increasing temperature the local stressed structure relaxes through breaking of Ge–Ge bonds (constituting the ethane-like unit) and acquisition of a nearby Se atom (belonging to Se–Se wrong bonds) to construct corner-shared Ge(Se 1/2) 4-tetrahedral units, and thereafter various crystalline phases of GeSe 2 results depending on the temperature.

Energetic Calculation of Coincidence Grain Boundaries with a Modified Stillinger-Weber Potential

Potential energy of two coincidence grain boundaries, Σ = 7 and 19, was calculated with a Stillinger-Weber potential previously adapted to wurtzite GaN in order to take into account the Ga–Ga and N–N wrong bonds. Their atomic structures, determined before by high resolution electron microscopy, are based on 5/7 and 8 atom ring structural units. By using periodical boundary conditions, it is shown that the high angle grain boundary Σ = 7 (Θ = 21°79/[0001]) has always a smaller energy than Σ = 19 (Θ = 13°17/[0001]).

The atomic structure and properties of wurtzite GaN epitaxial layers

Structural investigations show that extended defects in GaN have multiple atomic configurations. The pure edge threading dislocations exhibit 5/7, 4 or 8 atom cycles along their line. The dislocations are now known to have a strong effect on the transport as well as on the optical properties of the layers. The basal stacking faults (I1, I2 and E) form mostly close or inside the buffer layer. They were shown to be associated with photoluminescence peak close to the conduction band minimum (3.425 eV). The prismatic stacking fault bears two atomic configurations 1/2 〈101̄1〉 and 1/6 〈202̄3〉 whose stability is different in GaN, AlN or InN. The {101̄1} inversion domains which cross the whole epitaxial layers occur in particular growth conditions, they exhibit two surface morphologies and ranges of size. The smaller ones (<20 nm) terminate on a flat surface, they contain wrong bonds inside the boundary planes. The large ones (50 nm) are pyramid topped and their boundaries are reconstructed.

Ab initio study of structure in boron nitride, aluminum nitride and mixed aluminum boron nitride amorphous alloys

Car–Parrinello molecular dynamics is applied to the liquid quench simulation of the amorphous networks of BN, AlN and AlBN2. Structural and elastic properties were determined. It is found that AlN has a stronger tendency to chemical order than BN, driven by the greater energy penalty for “wrong bonds.” AlN, however, has a stronger tendency to form an amorphous structure as judged by the energy difference between the crystalline and amorphous states. There is some experimental evidence for an amorphous form of AlN. BN was simulated at two densities, 2.0 and 3.0 g/cm3. Even at the higher density, the fraction of tetrahedral coordination remained low, in contrast to AlN, enabling us to predict that the tetrahedral amorphous form of BN does not form under liquid quench conditions. The amorphous network with the formula AlBN2 has an intermediate tendency to form a tetrahedral structure and has a relatively high elastic modulus. This material is predicted to be of value for application as a wear resistant coating.

Band gap engineering in amorphous AlxGa1−xN: Experiment and ab initio calculations

Amorphous alloys of aluminum nitride and gallium nitride deposited at 100 K at compositions ranging from pure AlN to pure GaN with optical band gaps which vary linearly with composition from 3.27 eV (a-GaN) to 5.95 eV (a-AlN) have been synthesized. Ab initio molecular dynamics calculations for these alloys reproduce the band gap versus composition data and give specific information on the electronic localization of the band tail states. There are no midgap states in amorphous AlxGa1−xN alloys. The calculated models have mixed four-fold and three-fold coordination and have no wrong (homopolar nuclear) bonds, demonstrating the strong ionicity in amorphous AlxGa1−xN alloys. It has been found that the valence band tail states are mostly localized on the three-fold coordinated N sites while the conduction band tail states are mostly localized on the three-fold coordinated Ga or Al sites.

An empirical potential for the calculation of the atomic structure of extended defects in wurtzite GaN

The Stillinger–Weber potential has been parametrized for GaN with bond-type dependent parameters to allow efficient atomic simulations of large systems containing wrong, dangling and extra bonds. The input data for the fit are the experimental elastic constants of wurtzite structure and wrong bond energies deduced from ab initio calculation. The potential in then applied to zincblende structure and to planar defects. The predicted values are compared to experimental observation and previous computations.

Effect of thermal annealing on dynamics of photoluminescence in a-GeSe 2 films

We have investigated the effect of thermal annealing on dynamics of photoluminescence (PL) in vacuum-evaporated a-GeSe 2 films with thickness of 15 μm. The intensity of PL increases over 200 times after annealing at 250°C, which makes it possible to take a first investigation on both decay time and fatigue effect. The disordered amorphous structure of the a-GeSe 2 film, containing wrong bonds, Ge–Ge and Se–Se, has been reconstructed to a stable structure by diminishing the number of the wrong bonds and the related strain or internal pressure by thermal annealing.

An improved neural classification network for the two-group problem

In this paper we present the neural network model known as the mixture-of-experts (MOE) and determine its accuracy and its robustness. We do this by comparing the classification accuracy of MOE, backpropagation neural network (BPN), Fisher’s discriminant analysis, logistics regression, k nearest neighbor, and the kernel density on five real-world two-group data sets. Our results lead to three major conclusions: (1) the MOE network architecture is more accurate than BPN; (2) MOE tends to be more accurate than the parametric and non-parametric methods investigated; (3) MOE is a far more robust classifier than the other methods for the two-group problem. Scope and purpose The two-group classification problem is the assignment of objects to one of two predetermined groups. These classification decisions are routinely made in business, health care, and government. Classification errors result when an object is assigned to the wrong group. A healthy patient might be diagnosed with cancer, or a patient suffering from cancer might be diagnosed as healthy. A firm assigned the wrong bond rating will incur additional cost of capital of millions of dollars. Unfortunately, there is no classification method that is the “best” for all data sets. Some classification methods will do well for a given data set only to perform poorly on another. What is needed, therefore, is a robust methodology that classifies accurately across a wide range of two-group data sets. Researchers and practitioners have turned to neural classification models in a quest for a robust method. The purpose of this research is to investigate the accuracy and robust nature of the mixture-of-experts neural network (MOE). Our research demonstrates that MOE performs very well against back-propagation neural networks and four traditional statistical classification procedures based upon results from five diverse real world data sets. We conclude that the mixture-of-experts neural network may be the robust classification scheme being sought for use with the two-group problem.

Ab Initio Tensile Tests of Grain Boundaries in SiC

Ab initio tensile tests, namely, first-principles calculations of the tensile strength and fracture have been applied to a coincidence boundary in SiC for the first time, using the first-principles molecular-dynamics method beased on the density-functional theory. The non-polar interface of the {122}Σ=9 boundary has been dealt with. The stable configuration contains reconstructed five-membered and seven-membered rings and interfacial C-C and Si-Si wrong bonds. This interface is very strong because of the reconstruction, although the C-C and Si-Si bonds have large effects. The back bond of the C-C bond is broken first because of the short bond length and large strength of the C-C bond. Then the interfacial Si-C bonds are broken, and the Si-Si bond is broken. The ab initio tensile test is a powerful tool to investigate the intrinsic strength and fracture of interfaces through the behavior of electrons.

Theory of the (12_10) prismatic stacking fault in GaN

The atomic and electronic structure for a (12_10) prismatic stacking fault in GaN is calculated from first principles. The fault vector is 1/2[101_1]. Such a fault is believed to occur when GaN is grown on sapphire or SiC, and may originate when basal plane stacking faults fold onto the prism plane. We find that the boundary is heavily reconstructed, but that the atoms remain fourfold coordinated, with no wrong bonds or dangling bonds present. Despite the existence of distorted four-membered rings of bonds at the boundary, the fault does not induce deep states inside the gap. The calculated stacking fault energy is 72 meV/Å2.

Structural, electronic, and dynamical properties of amorphous gallium arsenide: A comparison between two topological models

We present a detailed study of the effect of local chemical ordering on the structural, electronic, and dynamical properties of amorphous gallium arsenide. Using the recently-proposed ``activation-relaxation technique'' and empirical potentials, we have constructed two 216-atom tetrahedral continuous random networks with different topological properties, which were further relaxed using tight-binding molecular dynamics. The first network corresponds to the traditional, amorphous, Polk-type, network, randomly decorated with Ga and As atoms. The second is an amorphous structure with a minimum of wrong (homopolar) bonds, and therefore a minimum of odd-membered atomic rings, and thus corresponds to the Connell-Temkin model. By comparing the structural, electronic, and dynamical properties of these two models, we show that the Connell-Temkin network is energetically favored over Polk, but that most properties are little affected by the differences in topology. We conclude that most indirect experimental evidence for the presence (or absence) of wrong bonds is much weaker than previously believed and that only direct structural measurements, i.e., of such quantities as partial radial distribution functions, can provide quantitative information on these defects in a-GaAs.

The atomic structure of {101̄0} inversion domain boundaries in GaN/sapphire layers

Using high resolution electron microscopy and extensive image simulation, the atomic structure of the inversion boundaries has been determined in GaN layers grown on sapphire by electron cyclotron resonance assisted molecular beam epitaxy. They form nanometric domains (5–20 nm) limited by {101̄0} planes crossing the whole epitaxial layer. These small dimensions excluded the use of more conventional microscopy methods such as convergent beam diffraction for their characterization. For image simulations, up to 10 models including the well known Austerman–Gehman and Holt ones were considered for the boundary atomic structure. The overall agreement with the observed contrast was reached for a Holt type model of structure in which the boundary plane contains two wrong bonds per atom. Although the overall atomic configuration is neutral in {101̄0} boundary planes, it is expected to be highly energetic in ionic materials like GaN due to the difference in Ga–Ga and N–N bond length and local charge difference. This may probably explain the small size of the domains observed in the present samples.

Theoretical studies of grain boundaries in covalent materials

Atomic and electronic structures of grain boundaries in covalent materials such as semiconductors or covalent ceramics have been investigated using various theoretical methods. Significant advances have been made in the understanding of grain boundaries in elemental semiconductors. Currently, ab initio calculations for grain boundaries are possible by virtue of the development of the first-principles molecular dynamics method. A first ab initio approach to grain boundaries in SiC by using the models of the {122} Σ = 9 boundary is presented. It has been shown that interfacial C-C and Si-Si wrong bonds have significant effects on grain boundaries in SiC. The C-C and Si-Si bonds have bond lengths and bond charges similar to those in bulk diamond and Si, and the C-C bonds generate peculiar localized states at the valence-band edges.

A structural study of gallium lanthanum sulphide glass bulk and thin films by x-ray absorption fine structure spectroscopy

Atomic-scale structure changes in gallium lanthanum sulphide bulk glass and ablation-deposited thin films have been studied by the x-ray absorption fine-structure (EXAFS) technique. EXAFS spectra have been recorded at the sulphur and gallium K edges, and the lanthanum edge, and this has allowed us to construct a detailed picture of the local structure in bulk glass and thin films. The EXAFS results indicate that there is chemical disorder in the structural network of the GLS thin films, although chemical ordering is predominant in bulk GLS glass. The existence of `wrong bonds', i.e. Ga - Ga and S - S bonds, in the structure has been discussed and correlated with optical absorption experiments undertaken on the same samples to provide a consistent picture of the local structure.

Topology of Amorphous Tetrahedral Semiconductors on Intermediate Length Scales

Using the recently-proposed ``activation-relaxation technique'' for optimizing complex structures, we develop a structural model appropriate to a-GaAs which is almost free of odd-membered rings, i.e., wrong bonds, and possesses an almost perfect coordination of four. The model is found to be superior to structures obtained from much more computer-intensive tight-binding or quantum molecular-dynamics simulations. For the elemental system a-Si, where wrong bonds do not exist, the cost in elastic energy for removing odd-membered rings is such that the traditional continuous-random network is appropriate. Our study thus provides, for the first time, direct information on the nature of intermediate-range topology in amorphous tetrahedral semiconductors.

The Atomic Structure of the {1010} Inversion Domains in GaN/Sapphire Layers

ABSTRACTNanometric inversion domains in GaN/Al2O3 layers have been investigated using HREM. They were found to be limited by {1010} planes and to cross the entire epitaxial layer. It has been possible, using extensive image simulation and matching to discriminate between possible atomic models for the boundary plane. It is shown that the inversion domain boundaries correspond to a Holt type model containing wrong bonds (Ga-Ga, N-N), and in that plane, each atom exhibits two such bonds. This probably can explain the small size of the domains (5–20 nm).

Structural and Electronic Properties of a-Gaas: A Tight-Binding–Molecular-Dynamics–Art Simulation

ABSTRACTBy combining tight-binding (TB) molecular dynamics (MD) with the recently-proposed activation-relaxation technique (ART), we have constructed structural models of a-GaAs and a-Si of an unprecedented level of quality: the models are almost perfectly four-fold coordinated and, in the case of a-GaAs, exhibit a remarkably low density of homopolar bonds. In particular, the models are superior to structures obtained using melt-and-quench TB-MD or quantum MD. We find that a-Si is best described by a Polk-type model, while a-GaAs resembles closely the mechanical model proposed by Connell and Temkin, which is free of wrong bonds. In this paper, the structural, electronic, and dynamical properties of a-GaAs based on this approach will be reviewed, and compared to experiment and other structural models. Our study provides much-needed information on the intermediate-range topology of amorphous tetrahedral semiconductors; in particular, we will see that the differences between the Polk and Connell-Temkin models, while real, are difficult to extract from experiment, thus emphasising the need for realistic computer models.