Chen Nianyi

Artificial neural network prediction of the band gap and melting point of binary and ternary compound semiconductors

Abstract In this paper, an artificial neural network trained by experimental data has been used to predict the values of the band gap and melting point of III–V, II–VI binary and I–III–VI 2 , II–IV–V 2 ternary compound semiconductors. The calculated results were in good agreement with the experimental ones.

Effect of Rare Earth Compositions on the Hydrogen Storage Properties of AB 5‐Type Alloys and Pattern Recognition Analysis

Among functional materials, hydrogen storage alloys have attracted much attention for more than a decade. AB{sub 5}-type alloys (La-Ce-Pr-Nd)(Ni-Mn-Al-Co){sub 5} have been explored for possible use as a negative electrode in Ni-metal hydrogen batteries. The initial hydrogen capacities and capacity ratios after charge-discharge cycles for AB{sub 5}-type alloys with different atomic ratios in La, Ce, Pr, and Nd are analyzed by the pattern recognition method, and some criteria for improving their hydrogen storage properties are obtained.

Calculation of thermodynamic properties from the miscibility gap in the phase diagram of Zn-Pb system by means of NRTL equation

Abstract In this paper, the NRTL equation approach has been used for a calculation of the activity coefficients of the components in a binary alloy system showing liquid phase immiscibility. The parameters needed for the calculation are the energy parameters, (g12–g22), (g21-g11) and the non-random parameter, α, which are determined by solving the NRTL equation with the aid of genetic algorithm. The calculation was carried out numerically for an immiscible phase in the binary alloy system, Zn-Pb. The agreement between the calculated and experimentally determined values of activity coefficient is excellent, indicating the validity of the NRTL equation model for the description of immiscible solutions.

Some regularities of melting points of AB-type intermetallic compounds

Abstract A modified cellular model is proposed for the investigation of melting point regularities of AB-type intermetallic compounds. By this model, five parameters of constituent elements: electronegativity difference, ΔX; valence electron density difference, Δ( Z R 3 ) ; electron-atom ratio, e a ; metallic radius ratio, R A R B ; and the average melting point of constituent elements, Tavg, are used to find a mathematical model for the melting point prediction, with the artificial neural network as the method of computation. The error of prediction is usually about 5% of the absolute temperature of melting point.

On the formation of ternary alloy phases

Abstract More than half of the ternary alloy phases have crystal structures derived from those of binary alloy phases as substitution derivatives. Atomic parameter diagrams and the pattern recognition method have been used for finding the formation criteria of ternary alloy phases of this type. Some empirical relationships found may be extended to study the formation of ternary alloy phases of other types.

Regularities of formation of ternary intermetallic compounds: Part 2. Ternary compounds between transition elements

Abstract Regularities of the ternary intermetallic compound formation between three transition elements are studied by the pattern recognition–atomic parameter method. The influences of VE, R sp and X MB on the formation of the ternary intermetallic compounds are investigated. The criteria of ternary compound formation in the late transition element-containing systems, early transition element-containing systems, systems containing both the early transition element and late transition element, and the systems containing Mn-group elements are found by pattern recognition methods.

Regularities of formation of ternary intermetallic compounds Part 4. Ternary intermetallic compounds between two nontransition elements and one transition element

Abstract Regularities in the formation of ternary intermetallic compounds consisting of one kind of transition element (T) and two kinds of nontransition elements (M, M′) have been studied by atomic parameter-pattern recognition methods. Some empirical criterions found are explained by the concepts of spd hybridization and charge transfer between the atoms of the constituent elements.

Computer-aided materials design of La-rich MH electrode materials

Abstract A partial least square–back propagation network (PLS–BPN) method is applied to the materials design of hydrogen storage alloys and property prediction of metal hydride (MH) electrodes. Two samples of hydrogen storage alloys designed from an optimal region of the PLS sub-space have been prepared and the predicted target values have been verified by experiment. Some properties of MH electrodes: initial capacity, capacity ratio and discharge curve predicted by the method are in agreement with experimental results.

Regularities of formation of ternary alloy phases between non-transition metals

Using a four-parameter model based on extended Miedema’s cellular model of alloy phases and pattern recognition methods, the regularities of formation of ternary intermetallic compounds between non-transition metals have been investigated. The criterion of formation can be expressed as some empirical functions of ϕ (electronegativity),n WS 1/3 (valence electron density in Wagner-Seitz cell),R (Pauling’s metallic radius) andZ (number of valence electrons in atom).Using a four-parameter model based on extended Miedema’s cellular model of alloy phases and pattern recognition methods, the regularities of formation of ternary intermetallic compounds between non-transition metals have been investigated. The criterion of formation can be expressed as some empirical functions of ϕ (electronegativity),nWS1/3 (valence electron density in Wagner-Seitz cell),R (Pauling’s metallic radius) andZ (number of valence electrons in atom).

Regularities of melting points and melting types of simple and complex ionic solids

The melting points and melting types (congruent or incongruent melting) of simple or complex ionic compounds are investigated by using artificial neural network with some atomic parameters (ionic radii, valency and electronegativity of constituent elements) as inputs. The artificial neural network trained by data of known compounds of the same valence type can be used to predict the melting type and melting points of other ionic compounds. The error of the results of computerized prediction is usually less than 5%.