Initial Cluster Size Research Articles

Laser modification of silver nanoclusters in SiO2 thin films

Thin films of silica containing silver nanoclusters have been deposited by magnetron co-sputtering followed by thermal annealing. Laser modification of the mean cluster size was performed using the fourth harmonic of a Nd:YAG laser with energies of between 35 and 125 mJ/cm2. The mean size of the clusters was estimated from the shape of the plasmon resonance band in the optical absorption spectra with the help of a computer simulation based on the Mie theory in static approximation. It was found that laser treatment with fluences above a certain threshold leads to a reduction of the mean size of the clusters and this reduction is greater for greater fluences. After a long treatment with the same fluence the effect saturates. The final mean size of the clusters after saturation depends only on the laser fluence and not on the initial mean cluster size. When lower laser fluences were used it was possible after laser annealing to return the mean cluster size to its initial value by thermal annealing. In this way by using a combination of laser treatment and thermal annealing a predetermined mean cluster size could be achieved. The mechanism of laser-induced cluster-size modification is discussed.

Shear-induced aggregation and break up of fibril clusters close to the percolation concentration

To probe the behaviour of fibrillar assemblies of ovalbumin under oscillatory shear, close to the percolation concentration, cp (7.5%), rheo-optical measurements and Fourier transform rheology were performed. Different results were found close to cp (7.3%), compared to slightly further away from cp (6.9 and 7.1%). For 6.9 and 7.1%, a decrease in complex viscosity, and a linear increase in birefringence, Δn′, with increasing strain was observed, indicating deformation and orientation of the fibril clusters. For 7.3%, a decrease in complex viscosity was followed by an increase in complex viscosity with increasing strain, which coincided with a strong increase in Δn′, dichroism, Δn″, and the intensity of the normalized third harmonic (I3/I1). This regime was followed by a second decrease in complex viscosity, where Δn′,Δn″ and I3/I1 decreased. In the first regime where the viscosity was decreasing with increasing strain, deformation and orientation of existing clusters takes place. At higher oscillatory shear, a larger deformation occurs and larger structures are formed, which is most likely aggregation of the clusters. Finally, at even higher strains, the clusters break up again. An increase in complex viscosity, Δn′, Δn″ and I3/I1 was observed when a second strain sweep was performed 30 min after the first. This indicates that the shear-induced cluster formation and break up are not completely reversible, and the initial cluster size distribution is not recovered after cessation of flow.

개선된 밀도 기반의 퍼지 C-Means 알고리즘을 이용한 클러스터 합병

1960년대 퍼지 이론이 소개된 이후 데이터 마이닝을 포함한 기계 학습 분야의 군집화 작업에서 퍼지 이론이 폭넓게 사용되었다. 퍼지 C-평균 알고리즘은 가장 많이 사용되는 퍼지 군집화 알고리즘이다. 이 알고리즘은 하나의 데이터 개체가 서로 다른 소속 정도를 가지고 각 군집에 할당될 수 있도록 한다. 퍼지 C-평균 알고리즘도 K-평균 알고리즘과 같은 일반적인 군집화 알고리즘과 마찬가지로 초기 군집수와 군집 중심의 위치에 의해 최종 군집 결과의 성능 차이가 나타난다. 군집화를 위한 이러한 초기 설정은 주관적이며 이 때문에 적절치 못한 결과를 얻게 될 수도 있다. 본 논문에서는 이 문제를 해결할 수 있는 방법으로 주어진 학습 데이터의 속성을 기반으로 한 초기 군집수와 군집 중심을 결정하는 개선된 밀도 기반의 퍼지 C-평균 알고리즘을 제안하였다. 제안 방법은 격자를 사용하여 초기 군집 중심의 위치와 군집수를 결정하였다. 기존에 많이 이용되었던 객관적인 기계 학습 데이터를 이용하여 제안 알고리즘의 성능비교를 수행하였다. The fuzzy set theory has been wide used in clustering of machine learning with data mining since fuzzy theory has been introduced in 1960s. In particular, fuzzy C-means algorithm is a popular fuzzy clustering algorithm up to date. An element is assigned to any cluster with each membership value using fuzzy C-means algorithm. This algorithm is affected from the location of initial cluster center and the proper cluster size like a general clustering algorithm as K-means algorithm. This setting up for initial clustering is subjective. So, we get improper results according to circumstances. In this paper, we propose a cluster merging using enhanced density based fuzzy C-means clustering algorithm for solving this problem. Our algorithm determines initial cluster size and center using the properties of training data. Proposed algorithm uses grid for deciding initial cluster center and size. For experiments, objective machine learning data are used for performance comparison between our algorithm and others.

붓스트랩 기법과 유전자 알고리즘을 이용한 최적 군집 수 결정

Optimal determination of cluster size has an effect on the result of clustering. In K-means algorithm, the difference of clustering performance is large by initial K. But the initial cluster size is determined by prior knowledge or subjectivity in most clustering process. This subjective determination may not be optimal. In this Paper, the genetic algorithm based optimal determination approach of cluster size is proposed for automatic determination of cluster size and performance upgrading of its result. The initial population based on attribution is generated for searching optimal cluster size. The fitness value is defined the inverse of dissimilarity summation. So this is converged to upgraded total performance. The mutation operation is used for local minima problem. Finally, the re-sampling of bootstrapping is used for computational time cost.

Spillover effects and the evolution of firm clusters

In this paper we consider the impact of spillovers occurring within each of two groups of firms on the long run agglomeration patterns in a market. In each period every single firm can either produce for this market or choose some outside option (e.g. a risky asset). Firms switch between the two options based on information about the relative profitability of the market and the outside option. In the market, due to spillovers, the production costs are influenced by the number of firms from the same population which are in the market. The resulting model describes the evolution of the size of the two firm clusters and their market shares over time. We provide a global analysis of the existence and basins of attraction of equilibria to address the question what impact different constellations of spillover effects have on the growth of dominant respectively incoming clusters. We demonstrate that the basins of attraction of coexisting long run equilibria do not depend continuously on the size of the spillover effects. Furthermore, an increase in the initial cluster size is not necessarily beneficial if the switching behavior of firms is fast.

A Laser Technique to Quantify the Size, Porosity, and Density of Clay Clusters During Sedimentation

Abstract A technique is developed for measuring the size, porosity, and density of clay clusters (flocs) and the distributions of these fundamental quantities during sedimentation. A laser-particle counter is used to detect cluster size and size distribution. Cluster velocities are computed by taking measurements at successive time intervals. By comparing the initial cluster size distribution with the measured distribution as sedimentation proceeds, a unique relationship between cluster size and settling velocity can be established. By applying a settling model for porous spheres, a unique relationship between cluster diameter and porosity is then established. Cluster density is calculated using fundamental mass-volume relationships. The technique and its working principles are illustrated via two sets of experiments, one for kaolinite clusters in distilled water and another for kaolinite clusters in 0.5 M NaCl. Consistent with double-layer theory, the results show that the porosity of clusters in distilled water is higher than the porosity of clusters in NaCl solution. The results also show that power-law relationships, which state that cluster porosity increases with cluster size, may not be valid for clusters very close to the primary particle size.

Approaching bulk limit for three-dimensional solids via the cyclic cluster approximation: Semiempirical INDO study

The cyclic cluster method has been examined for a number of solids using a recently developed computer code, Solid98. Calculations are based on the quasirelativistic (QR) INDO/1 (intermediate neglect of differential overlap) method that is simple enough to allow for a saturation of the (cyclic) clusters. Convergence toward the bulk limit (INDO/1) charge density with respect to the size of the model cyclic cluster is shown for diamond, silicon, germanium, boron nitride, gallium phosphide, gallium arsenide, and gallium antimonide. Results show that, as soon as the initial cluster size reaches 5 to 6 nm, one can safely use the obtained density matrix as a good approximation to the bulk limit. ©1999 John Wiley & Sons, Inc. J Comput Chem 20: 253–261, 1999

Influence of sub-surface defects on low-cycle fatigue life in a gas turbine disc alloy at elevated temperature

The influence on fatigue life of sub-surface non-metallic inclusions has been studied in Waspaloy at 650°C under both load and effective strain control at a strain ratio of R=0.1. The specimens were shot peened to represent the component condition and to suppress free surface initiation. When compared with defect-free material in terms of total strain range, the presence of the defects reduced average cyclic lives by an order of magnitude. Use of a strain-life based model incorporating a mean stress effect, e.g. the Smith, Watson and Topper approach, could be used for life prediction, but this would require a large data base and/or large reserve factor on life. Alternatively, the use of the initial inclusion cluster size and LEFM shows a reasonable correlation in terms of the effective stress intensity range (Δ K eff) and life, suggesting that confident safe life prediction requires the quantification of behaviour in terms of crack development and growth from the defects.

Charge transfer and fragmentation of liquid helium clusters that contain one or more neon atoms

An investigation of the electron impact ionization and fragmentation of helium clusters that contain Ne atoms and Nek subclusters has been performed. The charge transfer probability from He+ to Ne and the branching ratios for fragmentation of the Nek subclusters were found by analyzing the dependence of the ion signal intensities on the Ne pressure in the “pickup” region. The measured charge transfer probability from He+ to Ne ranges from 0.06±0.01 for clusters of mean original size 〈N〉=3300 to 0.43±0.02 for 〈N〉=1100. Charge transfer to a single Ne atom within the helium clusters never yields bare Ne+ ions. Instead, fragments of the type NeHen+ are produced. The charge transfer from He+ to Ne2 subclusters yields mainly Ne2+ for smaller initial cluster sizes, but NeHen+ or Ne2Hen+ fragments are more probable for larger clusters. This shows that He droplets of a few thousand atoms are able to cage Ne2 subclusters by dissipating the entire energy released by charge transfer and formation and vibrational relaxation of the Ne2+ ion. Interestingly, it was found that in these relatively small helium clusters the Ne3 and Ne4 subclusters never survive the charge transfer from He+. Fragments such as Ne2+ and Ne2Hen+ are more likely to survive than are Ne3+ and Ne4+. In general, the results presented here are qualitatively similar to those for a recent study of the ionization of Ar in helium droplets. In both cases fragmentation to the bare ion is rare, while fragmentation to the dimer ion dominates. However, the helium cluster caging effect is more efficient for Ne subclusters than for Ar subclusters. Also, there is no evidence for shell structures in the NeHen+ ion fragment distributions.

Capture and ionization of argon within liquid helium droplets

Liquid helium droplets of initial mean cluster size, 〈N〉, ranging from 600 to 8000 atoms are doped with argon using the pick-up technique. The doped clusters are ionized by electron impact, and the resulting fragment ions are monitored as a function of argon pressure in the pick-up volume. Analysis of the pressure dependent ion signals is used to determine (1) the probability for charge transfer from He+ to the Ar atoms within the droplet, and (2) the probability for fragmentation of the Ark subclusters upon ionization. The measured charge transfer probability from He+ to Ar ranges from 0.05±0.02 for clusters of mean original size 〈N〉=8000 to 0.26±0.05 for 〈N〉=600. Charge transfer to the Ark constituent results in the following qualitative trends; a single Ar atom yields HenAr+ ions; Ar2 mainly yields Ar2+, and Ar3 mainly fragments to yield Ar2+. Simulations of the results are performed to extract information on how the charge transfer and fragmentation processes within the ionized droplet dependent on the size of the helium droplet and the number of argon atoms captured. We use the positive-hole resonant-hopping mechanism to determine that the He+ hops 3–4 times prior to localization with either the Ar dopant or another He atom to form He2+. This corresponds to a time scale for He2+ formation of 60–80 fs.

Precrystallization structures in supersaturated lysozyme solutions studied by dynamic light scattering and scanning force microscopy

A comparitive study of the nanostructure evolving during aggregation of hen-egg white lysozyme in supersaturated solution was carried out by dynamic light scattering (DLS) and scanning force microscopy (SFM). Lysozyme aggregate (cluster) formation was observed in solution in the presence of NaCl, (NH4)2SO4, and NaNO3 as precipitating agents. The growth kinetics were examined by DLS and revealed fractal growth of the clusters with a fractal dimension of 1.8 obtained independently of the type of inert salt. Such behavior is typical for diffusion-limited cluster–cluster (DLCA) aggregation. Initial lysozyme cluster sizes were in the range of 12–35 nm. SFM images of individual lysozyme clusters at the liquid–solid interface were obtained in the presence of NaCl and NaNO3 under crystallization conditions, and revealed cluster sizes in agreement with those determined by DLS. Extended domains of smaller sized clusters appeared on the mica surface after subjecting supersaturated lysozyme solutions to a dialysis step. The feasibility of DLS and SFM for monitoring the nano- and mesoscopic morphology of lysozyme aggregates in supersatured solutions and at the solid–liquid interface is discussed.

Survival of noble gas clusters scattering from hot metal surfaces

We present classical trajectory calculations of Ar n ( n = 1000−400) colliding with a hot Pt(111) surface. Large cluster fragments are found to survive a surface collision, and the fraction of atoms remaining in the fragment is concluded to increase with initial cluster size and surface temperature, and decrease with incident velocity above 100 m/s. Up to 52% of the initial cluster is found to survive as one unit in the most favorable case of Ar 4000 scattering from a surface at 1500 K. The implications of the results for new experimental investigations are discussed.

A MODEL OF THE EVAPORATION OF BINARY-FUEL CLUSTERS OF DROPS

A formulation has been developed to describe the evaporation of dense or dilute clusters of binary-fuel drops. The binary fuel is assumed to be made of a solute and a solvent whose volatility is much lower than that of the solute. Convective flow effects, inducing a circulatory motion inside the drops, are taken into account, as well as turbulence external to the cluster volume. Results obtained with this model show that, similar to the conclusions for single, isolated drops, the evaporation of the volatile is controlled by liquid mass diffusion when the cluster is dilute. In contrast, when the cluster is dense, the evaporation of the volatile is controlled by surface layer stripping, that is, by the regression rate of the drop, which is, in fact, controlled by the evaporation rate of the solvent. These conclusions are in agreement with existing experimental observations. Parametric studies show that these conclusions remain valid with changes in ambient temperature, initial slip velocity between drops and gas, initial drop size, initial cluster size, initial liquid mass fraction of the solute, and various combinations of solvent and solute. The implications of these results for computationally intensive combustor calculations are discussed.

Population dynamics of Luehdorfia japonica Leech (Lepidoptera: Papilionidae)

SummaryMortality processes from egg to final instar larval stage are examined in the papilionid butterfly Luehdorfia japonica for two generations in a natural populations. Special attention is given to the effects of initial egg cluster size on the survival rate by the 3rd instar. Mean egg cluster sizes for the two generations were 11.3 and 10.7. The hatching rate was not affected by the egg cluster size. Up to 3rd instar, larvae in a group were more likely to die en masse rather than to die individually, whereas most larvae in later instars died independently. The egg clusters of the average (and the most frequent) size class (11±1) had the second highest survival rate until the 3rd instar. A slightly higher survival rate occurred in clusters a little larger (14±1). The average size clusters were least likely to be exterminated and about 70% of them produced at least one individual surviving to the 3rd instar. Factors affecting the observed mortality patterns were discussed.

Molecular-dynamics simulation of cluster and atom deposition on silicon (111).

The deposition of Si atoms and Si-atom clusters on a silicon (111) surface has been studied with molecular-dynamics simulations using two- and three-body interatomic Si potentials. The energy deposition and dynamics of Si-atom deposition have been studied as a function of substrate temperature, incident atom energy, and substrate thickness. Epitaxial Si(111) layers have been grown by single-atom deposition. The energy deposition and spreading of Si-atom clusters is characterized as a function of substrate temperature, initial cluster velocity, cluster temperature, and cluster size. A range of deposition parameters has been characterized that lead to high spreading of clusters, a feature that is necessary for achieving epitaxial thin-film growth.

Percolation model for simulation of char oxidation and fragmentation time-histories

A simulation model of char oxidation is introduced in which char particles are represented by randomly generated clusters of bonds on a lattice. Realizations are generated by means of a percolation algorithm. Bonds in contact with the ambience burn at a mean rate which depends on cluster size. By adjusting parameters governing this size dependence and the initial cluster size, simulated burning-rate time-histories are obtained which reproduce previously reported features of single-particle time-histories measured experimentally. The principal features are sharp increases in the burning rate, previously interpreted and here modeled as fragmentation events. Fragmentation is modeled as the separation of a cluster into disconnected pieces due to bond burnout. The agreement with experimental data lends support to this “percolative fragmentation” mechanism. The computations indicate the importance of the fragmentation process with regard to both its impact on the average burning time and its contribution to the production of fine particles.

Comparison of some different approximations in the statistical theory of relative dispersion

AbstractThe effect on the theory of relative dispersion of some different approximations to the two‐particle Lagrangian correlation function is examined. Two of these, one due to Taylor, the other to Smith and Hay, are treated in detail.Through numerical solution of the dispersion equations, the influence of the initial cluster size, a number of simple variations of the spatial separation argument of the two‐particle correlation function, the number of particles in the cluster and the ratio of Lagrangian to Eulerian integral scales are examined. With the exception of the initial cluster size, which is important in the early stage of growth, these parameters are relatively unimportant, particularly compared to the overall difference between the two approximations.In qualitative agreement with Batchelor's inertial range theory, both the Taylor and Smith‐Hay approximations show linear growth at small time with an accelerated growth region at intermediate time. However, between these regions, the Smith‐Hay solution shows a region of less‐than‐linear growth for which there appears to be no observational or theoretical support. This regime is more pronounced, and the difference between the two approximations greater, for initially small clusters.Comparison with suitably documented observations, while not entirely definitive, shows a degree of consistency and suggests the Taylor approximation to be the more appropriate.