The August issue of the Journal of the American Ceramic Society includes a special section on sintering comprising papers from the International Conference on Sintering 2011, which was held the Republic of Korea about a year ago.
The section opens with a review article by ACerS member Jian Luo of Clemson University titled “Developing Interfacial Phase Diagrams for Applications in Activated Sintering and Beyond: Current Status and Future Directions” (doi: 10.111/j.1551-2916.2011.05059.x). Specifically, the article reviews activated solid-state sintering and the development of interfacial phase diagrams.
Activated solid-state sintering arises from the presence of minor impurities that can promote densification rates at temperatures much lower than the eutectic or peritectic temperatures of the bulk composition. For example, adding less than one percent of a transition metal to tungsten or molybdenum powders has been shown to accelerate the sintering rates at temperatures that are 60 percent to 85 percent of the bulk eutectic or peritectic temperatures. Similar observations of sub-solidus sintering have been reported in ceramic systems such as ZnO with Bi2O3 additions and Ce2O with CoO additions and others.
Knowledge of the phenomenon is not new. Going back to 1842, Michael Faraday noticed that two block of ice can fuse together below 0 C, or that snowballs can hold together below freezing. At the time, he posited that the ice block “sintering” resulted from a “possible stabilization of a quasi-water layer” on the ice surface at temperatures below freezing that is possible because of impurities in the water.
The formation of stable, quasi-liquid surface layers at below-bulk melt temperatures is known as “surface melting” or “premelting” in the sintering literature. Grain boundary premelting, or “structural disordering,” has been studied extensively by physicists and more recently by materials scientists in unary systems, but its importance remains a point of controversy.
In multicomponent systems, however, complex intergranular phases have been observed and have been named “complexions.” The paper describes complexions as “a series of discrete GB phases, namely, an intrinsic (nominally “clean”) GB, a monolayer, a bilayer, a trilayer, a nanoscale intergranular film (IGF) of an equilibrium thickness and a complete wetting film of an arbitrary thickness.”
Luo outlines the key characteristics of impurity-based IGFs.
1. They have an equilibrium (self-selected) thickness on the scale of one nanometer,
2. The average film composition is different from the corresponding bulk liquid/glass phase,
3. The structure is neither fully crystalline, nor fully amorphous.
These nanoscale interfacial phases and their stability are of paramount importance. For example, intergranular film controls the room temperature toughness and high temperature creep resistance of silicon nitride. Similarly, crystallization of IGFs in silicon carbide trades off fracture toughness for improved creep resistance. Additional reports connect IGF and other grain boundary phases to erosive wear resistance in alumina, superplasticity in YSZ and embrittlement in refractory metals and cermets.
Luo observes that IGFs and complexions can be controlled, too, to tune the properties of functional materials. He cites examples like conductivity of ruthenate thick-film resistors, nonlinear I-V characteristics in ZnO varistors doped with Bi2O3 and thermal conductivity of aluminum nitride.
This may have a familiar sound to it. Several of the grand challenges of ceramic science determined in last spring’s workshop identify interfaces as critical. For example, functionality of oxide electronics is a question of designing surfaces, interfaces and nanoscale structure. Metastable defects near interfaces is another grand challenge (How do interfaces, which may be the dominant structural feature of nanoscale materials, and point defects interact?). Still another of the grand challenges simply asks whether the concept of an interface needs redefining (Can we “go beyond boundaries?”).
There needs to be a way to understand IGF behavior, however, and phase diagrams are the obvious tool. However, bulk diagrams do not apply. Luo reviews the emergence of CalPhaD methods to develop interfacial phase diagrams, the so-called λ-diagrams.
λ is a thermodynamic parameter that “represents the thermodynamic tendency for an interface to disorder and scales the actual (equilibrium) film thickness.” That is, it accounts for the thermodynamics, but also the effects of size and interfaces, which thermodynamic treatments usually ignore.
The article reviews in detail the thermodynamic equations behind calculating λ-diagrams, and then cites experimental examples that validate the approach. He shows a striking example of zinc oxide doped with Bi2O3 that was sintered below the ZnO solidus. A HRTEM micrograph shows a Bi2O3-rich, quasi-liquid film in most of the grain boundaries and on some of the (non-GB) surfaces.
Validation of the computational model for generating λ-diagrams opens the door to new approaches to sintering, what Luo calls “mechanism-informed materials design.” (Anyone else thinking Materials Genome Initiative?) He proposes a few examples.
• Predicting grain boundary-controlled high-temperature properties such as creep, oxidation and corrosion resistance;
• Designing fabrication protocols that yield interfacial structures that in turn optimize microstructures and morphologies; and
• Designing dopant and heat treatment approaches to control interfaces and, thus, properties.
Luo is frank and admits that λ-diagrams are not rigorous phase diagrams, but he feels they have a role in the toolbox for helping to predict trends. In his concluding remarks he says,
In general, understanding the stability of these interfacial phases and development of interfacial phase diagrams can help to control the mechanical and physical properties of … conventional materials as well as develop new classes of “interfacial materials” to achieve superior properties that are not attainable by bulk phases.