The MSE Paradigm - Defining the Discipline

Now that you've considered a perhaps "mundane" material from the world around you and perhaps the impact it's had on your life, it might be a good thing to more precisely define what a material is. Defining a material will also allow us to consider how (and why) the discipline of Materials Science and Engineering (MSE) is distinct from other fields of science and engineering.

For the purposes of this class, we will define a material simply: as a solid(ish) substance that has (potential) technological relevance or application. We'll exclude anything liquid or gas and leave their behaviors to chemical engineers. Further, we'll exclude materials that can't be used for anything interesting: materials scientists and engineers are particularly interested in those materials which have useful applications. We want to learn about the materials so that we understand how we might use them! This means that something like solid helium - regardless of how fundamentally interesting it might be - isn't of interest to materials scientists because we can't use it in a conventional engineering application - it is only only stable at extremely low temperatures and high pressures. That's why condensed matter physicists study it.

Within MSE itself, scientists and researchers usually distinguish themselves:

• Materials scientists are interested in understanding the why and how of materials behavior. Why is this material strong? Why does this material conduct electricity well? How might I make this material stronger? How can I make this material conduct electricity better?
• Materials engineers also think about the why and the how, but they typically have specific goals in mind within an engineering context. For example: I need a material that will resist a specific load in highly corrosive conditions. What do I need to engineer in order to make my material perform in this way? Ultimately, will my engineered material serve well in the application?

In order for materials scientists and engineers to answer these questions, we utilize a single guiding paradigm that defines the scope of the field. This is the called the Materials Science Paradigm (Figure 1.4.1), and it highlights the four central ideas in the study and engineering of materials:

• Processing: How a material is made. This includes, for example, casting of iron, curing of epoxy, firing of clay, or lay-up of carbon fiber.
• Structure: How a material's atoms or molecules are assembled. What patterns or arrangements do the atoms or molecules have? What types of defects and imperfections are present in the material? Examples are crystal structure, phase boundaries, dislocation arrangements, polymer conformation, all covered later in this course.
• Properties: How a material behaves. If I act on a material, how does it respond? For example, if I apply a force to a material by pulling on it, how far does it stretch? Similarly, if I apply a thermal gradient on a material by placing it between a hot and a cold area, how does the material transfer heat? Properties can be measured and quantified. For example, a material may have a "yield strength", measured in units of MPa, which tells us how much force can be applied before it starts to permanently deform.
• Performance: How a material functions in an application. That is, how does the combination of properties of a material serve in a specific application? This is often somewhat qualitative and depends on a combination of properties: How well does a material serve as a cell phone screen? Is a material well suited for use as an light-absorbing layer in a solar cell? Does a material work well for an airplane wing's leading edge?

The critical feature of the MSE Paradigm is that the processing-structure-properties-performance relationship is linear and direct. Figure 1.4.1 shows this important directional relationship. In this diagram, one can see that the processing of the material leads its structure - we control processing to yield different structures. That structure dictates the properties of the material. The properties determine the performance.

Importantly, the ovals represent the linear connection between each of the facets of the paradigm. Processing and structure are linked directly. However, processing and properties are not. This communicates that processing cannot directly affect properties. Processing can only change structure, and it is the structure that determines properties. Similarly, structure cannot determine performance. Structure gives properties, and the selection of properties yields our performance.

This is all well and good - but how do we use the MSE Paradigm? In order to understand and engineer materials, we need to explore the connections within the paradigm itself. There's many ways to do this - listed at the MSE Tools of the Trade in Figure 1.4.1.

We can use theory to explore these relationships: for example, we may come up with a model for how electrons move through different a structures (i.e., the Drude Model. Modeling immensely complex behaviors with mathematical models derived from physical understanding of the system has been the traditional foundational to all science and engineering for centuries, and there are of course MSE-specific theories that we'll leverage to understand and predict materials behaviors.

We can experimentally (or empirically) characterize materials behaviors to elucidate their behavior. Measuring, for example, what happens to the electrical conductivity of silver as we mix it with copper will tell us a great deal about how the materials composition (processing) affects its structure (the assembly of silver and copper atoms), and in turn how that structure affects its properties (its electrical conductivity). Experimentation is another cornerstone of science and engineering, allowing us to formulate new theories or adapt old ones.

Computation is a tool that allows us to model materials using computer simulations. We'll use the power of computational models in this class to learn about materials phenomena (bonding, crystal structure, diffusion, defects). Like characterization/experimentation, computation can help us develop new theories and improve old ones. Indeed, sometimes materials are now engineered on computers with simulations before they're synthesized in the lab! (See Xiong et al for a discussion of the role of computation in materials engineering.)

Finally - one of the newest tools in the toolbox is founded in data science approaches such as artificial intelligence and machine learning. Perhaps not surprisingly, discovering new materials and engineering them into useful products is an immensely difficult task. It takes human expertise (PhD-level training and years in the field), ingenuity, and resources (time and money) to develop the next new impactful material. Machines can explore materials solutions in ways are impossible or impractical for a human mind to conceive. As artificial intelligence has the power to drive a paradigm shift in other fields (self-driving vehicles, logistics, etc.), it too can afford new opportunities in materials science. Pyzer-Knapp et al provide a discussion of the role of artificial intelligence in MSE here: Pyzer-Knapp et al, Npj Comput. Mater., 2022.

We use all of these tools in concert in all subdisciplines of MSE to explore the connections in the paradigm, gain better understanding of our materials, and engineer new ones. In this course, we'll use theory, empiricism, and computation to learn about materials - and someday we hope fold in data science techniques, but we aren't quite there yet. Figure 1.4.2 shows how each of these scientific "tools" has led to an acceleration of the speed, automation, and scale of materials science and engineering .