section epub:type=”chapter”> This chapter discusses the kinetic theory of diffusion. The rate of steady-state creep varies with temperature. έss = Ce-(Q/ṜT). This is an example of Arrhenius’s law. Like all good laws, it has great generality. Thus, it applies not only to the rate of creep, but also to the rate of oxidation, the rate of corrosion, the rate of diffusion, even to the rate at which bacteria multiply and milk turns sour. The chapter states that the rate of the process (creep, corrosion, diffusion, etc.) increases exponentially with temperature (or, equivalently, that the time for a given amount of creep, or of oxidation, decreases exponentially with temperature). If the rate of a process that follows Arrhenius’s law is plotted on a loge scale against 1/T, a straight line with a slope of –Q/Ṝ is obtained. The value of Q characterizes the process. The chapter discusses the origin of Arrhenius’s law and its application to diffusion. This is followed by the discussion of diffusion and Fick’s law and data for diffusion coefficients. The chapter concludes with a discussion of mechanisms of diffusion. We saw in the last chapter that the rate of steady-state creep, Here Q is the activation energy for creep (J mol–1), R is the universal gas constant (8.31 J mol–1 K–1) and T is the absolute temperature (K). This is an example of Arrhenius’s law. Like all good laws, it has great generality. Thus it applies not only to the rate of creep, but also to the rate of oxidation (Chapter 25), the rate of corrosion (Chapter 27), the rate of diffusion (this chapter), even to the rate at which bacteria multiply and milk turns sour. It states that the rate of the process (creep, corrosion, diffusion, etc.) increases exponentially with temperature (or, equivalently, that the time for a given amount of creep, or of oxidation, decreases exponentially with temperature) in the way shown in Figure 22.1. If the rate of a process that follows Arrhenius’s law is plotted on a loge scale against 1/T, a straight line with a slope of In this chapter, we discuss the origin of Arrhenius’s law and its application to diffusion. In the next, we examine how it is that the rate of diffusion determines that of creep. First, what do we mean by diffusion? A good way of demonstrating it is to take a glass Petri dish, and run a very shallow layer of water into it. Wait until the water is totally still, and then drop a very small crystal of potassium permanganate into it. The edge of the crystal will immediately begin to dissolve in the water, producing a dark pink ring of solution centered on the crystal—see Figure 22.3(a). The ring then begins to spread radially outward into the layer of water—initially very quickly, but progressively more slowly as time passes. Figures 22.3(a) to 22.3(d) are time-lapse photographs showing the ring spreading. This is caused by the movement of potassium permanganate ions by random exchanges with the water molecules. The ions move from regions where they are concentrated to regions where they are less concentrated—the ions diffuse down the concentration gradient. This behavior is described by Fick’s first law of diffusion: J is the number of ions diffusing down the concentration gradient per second per unit area; it is called the flux of ions (Figure 22.4). c is the concentration of ions in the water—the number of ions (in this case permanganate ions) per unit volume of solution. x is the distance measured along the flux lines, and D is the diffusion coefficient for permanganate ions in solution—it has units of m2 s–1. This diffusive behavior is not just limited to ions in water—it occurs in all liquids, and more remarkably, in all solids as well. As an example, in the alloy brass—a mixture of zinc in copper—zinc atoms diffuse through the solid copper in just the way that ions diffuse through water. Because the materials of engineering are mostly solids, we shall now confine ourselves to talking about diffusion in the solid state. Physically, diffusion occurs because atoms, even in a solid, are able to move—to jump from one atomic site to another. Figure 22.5 shows a solid in which there is a concentration gradient of black atoms: there are more to the left of the broken line than there are to the right. If atoms jump across the broken line at random, then there will be a net flux of black atoms to the right (simply because there are more on the left to jump), and, of course, a net flux of white atoms to the left. Fick’s law describes this. It is derived in the following way. The atoms in a solid vibrate, or oscillate, about their mean positions, with a frequency υ (typically about 1013 s–1). The crystal lattice defines these mean positions. At a temperature T, the average energy (kinetic plus potential) of a vibrating atom is 3kT where k is Boltzmann’s constant (1.38 × 10–23 J atom–1 K–1). But this is only the average energy. As atoms (or molecules) vibrate, they collide, and energy is continually transferred from one to another. Although the average energy is 3kT, at any instant, there is a certain probability that an atom has more or less than this. A very small fraction of the atoms have, at a given instant, much more—enough, in fact, to jump to a neighboring atom site. It can be shown from statistical mechanical theory that the probability, p, that an atom will have, at any instant, an energy Why is this relevant to the diffusion of zinc in copper? Imagine two adjacent lattice planes in the brass with two slightly different zinc concentrations, as shown in exaggerated form in Figure 22.6. Let us denote these two planes as A and B. Now for a zinc atom to diffuse from A to B, down the concentration gradient, it has to “squeeze” between the copper atoms (a simplified statement—but we shall elaborate on it in a moment). This is another way of saying: the zinc atom has to overcome an energy barrier of height q, as shown in Figure 22.6. Now, the number of zinc atoms in layer A is nA, so the number of zinc atoms that have enough energy to climb over the barrier from A to B at any instant is For these atoms actually to climb over the barrier from A to B, they must of course be moving in the correct direction. The number of times each zinc atom oscillates toward B is ≈ υ/6 s–1 (there are six possible directions in which the zinc atoms can move in three dimensions, only one of which is from A to B). Thus the number of atoms that actually jump from A to B per second is But, meanwhile, some zinc atoms jump back. If the number of zinc atoms in layer B is nB, the number of zinc atoms that can climb over the barrier from B to A per second is
Kinetic Theory of Diffusion
Publisher Summary
22.1 Introduction
22.2 Diffusion and Fick’s Law