Search

In this chapter, the critical factors affecting oxygen permeation through dual-phase membranes are discussed with a view to improving the permeability of dual-phase membranes. Based on the current knowledge in solid-state electrochemistry, two series of dual-phase membranes, made of ceria-based ionic conductors and Fe-based perovskite mixed conductors, are suggested as promising dual-phase materials. Oxygen exchange on the membrane surface, preparation methods, sintering temperature, and phase composition are all critical factors to affect the membrane performance and discussed with experiment demonstration. Other factors, such as the elemental composition of the phases, grain size, grain size ratio between the two phases, the conductivity of the mixed conducting phase, the microstructure, and thickness, all affect the oxygen transport process and deserve thorough investigation.

Except for pure oxygen production, MIEC membranes can be used in membrane reactors for natural gas conversion. However, pure oxygen production is the process with the greatest potential for commercialization of the MIEC membrane technology in the near future. Air Products & Chemicals Inc. (APCI) is the pioneer on the commercialization of MIEC membrane technology for oxygen production, leading a team that includes several companies, institutes, and universities. In this chapter, the great progresses achieved by the APCI team are introduced, including the development history, membrane module design, membrane sealing, and module performance. Additionally, the progresses achieved by Aachen University are also briefly introduced.

Preparation method has a significant influence on the performance of MIEC membranes, because the membranes derived from different methods have different microstructures, defects, impurities, etc. In this chapter, the experimental methods related to the MIEC membranes are introduced, including the methods for the synthesis of ceramic powders, for the fabrication of MIEC membranes, and for the performance testing and characterization. The advantages and disadvantages of the methods for preparing ceramic powders and fabricating MIEC membranes are shown in this chapter to facilitate selecting proper methods in experiments.

Mixed ionic–electronic conducting (MIEC) membranes have attracted considerable attention from the industrial and academic communities over the past 30 years. This is deemed as a next-generation technology for pure oxygen production and powerful catalytic membrane reactors with high process intensification and large energy savings. This chapter briefly presents the history of MIEC membrane and provides an introduction to membrane reactors, the permeation principle, and the types of MIEC membranes, as well as outlines the scope of this book.

Although the intermediate–low-temperature operation has many advantages in practical applications, the operation temperature of MIEC membranes is limited in the range of 700–1000 °C in literatures, because membranes have low oxygen permeation flux and low permeation stability when the temperature is lower than 700 °C. With the development of preparation techniques of MIEC membranes, the low oxygen permeation flux can be easily solved; however, it is still a challenge to overcome the degradation of oxygen permeation in the intermediate–low temperature range until we published several papers to disclose the degradation mechanism and the stabilization methods recently. Thus, in chapter, an introduction of our achievements is presented on the stabilization of oxygen permeation of MIEC membranes in the intermediate–low temperature range.

Besides oxygen separation from air, another important application of MIEC membranes is to act as membrane reactors to unite reactions and separation. The integration of reactions and separation has many advantages, thus has attracted much attention of the industrial and academic communities. The important reactions in MIEC membrane reactors include partial oxidation of natural gas to syngas, oxidation of coupling of methane, oxidation dehydrogenation of ethane to ethylene, water splitting for hydrogen production, etc. To safely operate the membrane reactors for oxidation reactions, the membrane materials need to be stable in both oxidative and reductive atmospheres. Thus, in this chapter, the most concerned topic, i.e., membrane materials, is introduced, and the factors influencing the stability of membrane materials are discussed. Additionally, the progresses on the membrane reactors for various oxidation reactions are briefly introduced, and the reaction–separation–reaction membrane reactors developed in the recent years are presented.

Perovskite-type MIEC oxides are the most widely investigated membrane materials since the pioneering works of Teraoka and coworkers in 1985. In the early stage, cobalt-based perovskite oxides attracted much attention of researchers due to their high oxygen permeability. Ba0.5Sr0.5Co0.8Fe0.2O3−δ and La0.6Sr0.4Co0.2Fe0.8O3−δ are the two materials which have been thoroughly investigated in the past decades. Subsequently, the stability of the materials was regarded more important than the permeability for a practical application, and then cobalt-free perovskite membranes were extensively investigated in the past several years. In this chapter, the development of the cobalt-based and cobalt-free perovskite membranes is introduced in detail, and the influence of doping on oxygen permeability and stability is concerned.

Point defects are the main defect type concerned in the field of MIEC membranes, because they are closely related to the motion of oxygen vacancies and holes in crystalline solids. In the first part of this chapter, the concept of point defects is introduced briefly, and then, the knowledge about point defects in MIEC oxides, defect association, and defect equilibrium are introduced. Grain boundary is mentioned as an important two-dimensional defect because of its potential influences on the mechanical, electric, magnetic, catalytic, and thermal properties of materials. In the second part, the concept of diffusion, the diffusion mechanism, and diffusion paths of oxygen ions in perovskite- and fluorite-type ionic conductors are presented. There are several types of diffusion coefficients derived from different measuring methods, and the difference and relationships between these diffusion coefficients are provided.

The ionic conductivity of perovskite-type MIEC membranes is several orders of magnitude lower than their electronic conductivity; however, the ambipolar conductivity determines the diffusion rate of oxygen ions in the membrane bulk. Therefore, the improvement of oxygen ionic conductivity is the key to enhance the oxygen permeability of an MIEC material. In this chapter, the common oxygen ionic conductors with fluorite and perovskite structures are introduced in detail, and other types of oxygen ionic conductors, such as Bi4V2O11, La2Mo2O9, and La10-xSi6O26+y, are briefly presented. The critical factors influencing the oxygen ionic conductivity are discussed for the perovskite-type oxygen ionic conductors and MIEC conductors. All the oxygen ionic conducting materials can be made into MIEC membranes for oxygen permeation as long as they are properly doped by elements with variable valence states or mixed with a secondary phase with electronic conduction. MIEC membranes are all operated at high temperatures, and thus, the special properties related to high temperature, such as cationic diffusion, kinetic demixing, thermal expansion, chemical expansion, and creep, are briefly introduced to show that several factors should be considered to design or select a practically applicable MIEC membrane material.

Although the principle of oxygen permeation through an MIEC membrane is simple, it is difficult to develop a proper mathematical equation to describe the processes including oxygen exchange reactions on gas–solid interfaces and ambipolar diffusion in membrane bulk. In this chapter, the classical Wagner equation is introduced firstly. Only the oxygen permeation through an MIEC membrane is dominated by the ambipolar diffusion in membrane bulk; the permeation process can be well described by the Wagner equation. However, most of membranes are limited by the joint control of interfacial exchange and bulk diffusion. Therefore, three permeation models were developed by researchers to include the influence of oxygen exchange reactions on the gas–solid interfaces. The derivation and applications of the three models are separately introduced in detail in this chapter.

To solve multi-class classification problems for large-scale datasets, the authors propose a coded output support vector machine (COSVM) by introducing the idea of information coding. The COSVM is built based on the support vector regression (SVR) machine that is implemented by the sequential minimal optimization (SMO) algorithm. The paper first introduces the soft e-tube SVR's basic principles, next gives the idea and procedure of the SMO algorithm, and then illustrates the COSVM's topology. For studying the parameters impact on the binary COSVM's performance, we perform two experiments with the Character Trajectories dataset, in which output labels are coded with the binary number system. And some useful results are obtained in these experiments. The final section gives a conclusion and further research ideas.

In the vocabulary of Chinese language, idioms are a distinctive part for its fixed constitution grammatically and metaphorical meaning semantically. This paper introduces the entries and fields selected for the Chinese Idiom Knowledge Base (CIKB) built by the Institute of Computational Linguistics at Peking University (ICL/PKU) as well as its construction methods, research and applications so far. As a new member of the Comprehensive Language Knowledge Base (CLKB) constructed by ICL, our idiom knowledge base represents the new development of CLKB and will play a major role in NLP tasks such as machine translation, computer-aided translation, cross-language search, linguistic research, teaching Chinese as a foreign language and even as an electronic tool for preserving this non-material Chinese cultural and historical heritage.

The authors propose a coded output support vector machine (COSVM) by introducing the idea of information coding to solve multi-class classification problems for large-scale datasets. The COSVM is built based on the support vector regression (SVR) machine that is implemented by the sequential minimal optimization (SMO) algorithm. The paper first introduces the soft e-tube SVR's basic principles, next shows the idea and procedure of the SMO algorithm, and then gives the idea and topology of the COSVM. To study a number system's (NS) impact on the COSVM's performance, three experiments are performed with the Character Trajectories dataset, in which output labels are coded with the natural NS, decimal NS, and binary NS, respectively. Some useful results are obtained in these experiments. The final section concludes the paper and gives some further research visions.

Management of requirements inconsistency is key to the development of trustworthy software systems. But at present, although there are a lot of work on this topic, most of them are limited in treating inconsistency at the syntactic level. We still lack a systematical method for managing requirements inconsistency at the semantic level. This paper first proposes a requirements refinement model, which suggests that interactions between software agents and their ambiences are essential to capture the semantics of requirements. We suppose that the real effect of these interactions is to make the states of entities in the ambiences changed. So, we explicitly represent requirements of a software agent as a set of state transition diagrams, each of which is for one entity in the ambiences. We argue that, based on this model, the mechanism to deal with the inconsistency at the semantic level. A domain ontology is used as an infrastructure to detect, diagnose and resolve the inconsistency.