Wulff construction

Theory

In 1878 Josiah Willard Gibbs proposed that a droplet or crystal will arrange itself such that its surface Gibbs free energy is minimized by assuming a shape of low surface energy. He defined the quantity

Here γ j represents the surface (Gibbs free) energy per unit area of the j th crystal face and O j is the area of said face. Δ G i represents the difference in energy between a real crystal composed of i molecules with a surface and a similar configuration of i molecules located inside an infinitely large crystal. This quantity is therefore the energy associated with the surface. The equilibrium shape of the crystal will then be that which minimizes the value of Δ G i .

In 1901 Russian scientist George Wulff stated (without proof) that the length of a vector drawn normal to a crystal face h j will be proportional to its surface energy γ j : h j = λ γ j . The vector h j is the "height" of the j th face, drawn from the center of the crystal to the face; for a spherical crystal this is simply the radius. This is known as the Gibbs-Wulff theorem.

In 1953 Herring gave a proof of the theorem and a method for determining the equilibrium shape of a crystal, consisting of two main exercises. To begin, a polar plot of surface energy as a function of orientation is made. This is known as the gamma plot and is usually denoted as γ ( n ^ ) , where n ^ denotes the surface normal, e.g., a particular crystal face. The second part is the Wulff construction itself in which the gamma plot is used to determine graphically which crystal faces will be present. It can be determined graphically by drawing lines from the origin to every point on the gamma plot. A plane perpendicular to the normal n ^ is drawn at each point where it intersects the gamma plot. The inner envelope of these planes forms the equilibrium shape of the crystal.

Proof

Various proofs of the theorem have been given by Hilton, Liebman, Laue, Herring, and a rather extensive treatment by Cerf. The following is after the method of R. F. Strickland-Constable. We begin with the surface energy for a crystal

which is the product of the surface energy per unit area times the area of each face, summed over all faces. This is minimized for a given volume when

We then consider a small change in shape for a constant volume

which can be written as

The second term must be zero, as it represents the change in volume [which does not make sense, as before the whole expression was supposed to be the change in volume] and we wish only to find the lowest surface energy at a constant volume (i.e., without adding or removing material). We are then given from above

and

which can be combined by a constant of proportionality as

The change in shape ( δ O j ) must be allowed to be arbitrary, which then requires that h j = λ γ j , which then proves Gibbs-Wulff Theorem.