Dielectrics, Polarization and the D Field

The Physical Picture: Polarization

Imagine a dielectric material as a collection of neutral atoms. In the absence of an external electric field, the positive nucleus and the electron cloud are centered on each other. When an external field is applied, the positive charges are pulled slightly in the direction of the field, while the negative charges are pulled slightly against it. This separation of charge centers creates an electric dipole moment for each atom.

When we look at the material on a macroscopic scale, we define the polarization vector \(\vec{P}\) as the net electric dipole moment per unit volume. If the material is uniform and the field is constant, \(\vec{P}\) is constant throughout the bulk. However, at the surface of the dielectric, the dipoles are not perfectly cancelled by their neighbors. This results in a net accumulation of charge on the surface, known as bound surface charge. Inside the material, if the polarization is uniform, the bound charges cancel out perfectly. But if the polarization varies with position, there is a net accumulation of bound charge within the volume as well. This is the physical origin of the "bound charge" that complicates the \(\vec{E}\) field calculation.

Bound Charge Densities

To quantify the effect of polarization, we derive the charge densities associated with the dipoles. Consider a small volume element within the dielectric. The net charge inside this volume due to polarization is determined by how much dipole moment flows out of the volume. Mathematically, this is expressed as the negative divergence of the polarization field.

This equation tells us that if the polarization is uniform (\(\nabla \cdot \vec{P} = 0\)), there is no bound volume charge. However, at the boundary of the dielectric, the polarization drops to zero outside the material. This discontinuity creates a surface charge density. The bound surface charge density is given by the component of the polarization vector normal to the surface.

Here, \(\hat{n}\) is the unit normal vector pointing outward from the dielectric surface. These two equations, \(\rho_b\) and \(\sigma_b\), allow us to calculate the total charge distribution if we know the polarization. However, since \(\vec{P}\) itself depends on \(\vec{E}\), solving for \(\vec{E}\) directly is often a circular and difficult problem.

Defining the D Field

To resolve this difficulty, we define a new field, the electric displacement field \(\vec{D}\). The definition is constructed specifically to absorb the polarization term. We define \(\vec{D}\) as the sum of the vacuum permittivity times the electric field plus the polarization.

This definition is not arbitrary; it is chosen so that the divergence of \(\vec{D}\) yields only the free charge density. Let us verify this by taking the divergence of both sides. Using the linearity of the divergence operator, we get \(\nabla \cdot \vec{D} = \varepsilon_0 (\nabla \cdot \vec{E}) + \nabla \cdot \vec{P}\). From Gauss's law for the total electric field, we know that \(\varepsilon_0 (\nabla \cdot \vec{E}) = \rho_{\text{total}} = \rho_f + \rho_b\), where \(\rho_f\) is the free charge density and \(\rho_b\) is the bound charge density. Substituting the expression for bound volume charge \(\rho_b = -\nabla \cdot \vec{P}\), we find that the \(\nabla \cdot \vec{P}\) terms cancel out exactly.

This is the differential form of Gauss's law for the \(\vec{D}\) field. It states that the source of the \(\vec{D}\) field is only the free charge, which is the charge we explicitly place on conductors or electrodes. This is a powerful simplification because \(\rho_f\) is usually known or easily calculated, whereas \(\rho_b\) is a complex function of the field.

Integral Form and Boundary Conditions

By applying the divergence theorem to the differential equation, we obtain the integral form of Gauss's law for \(\vec{D}\). This form is particularly useful for calculating fields in symmetric geometries, such as parallel plate capacitors or spherical dielectrics.

Here, \(Q_{f,\text{enc}}\) is the total free charge enclosed by the Gaussian surface. Notice that bound charges do not appear on the right-hand side. This means that when using a Gaussian surface that cuts through a dielectric, we only need to account for the free charges on the plates, ignoring the induced bound charges on the dielectric surfaces.

At the interface between two different dielectric materials, or between a dielectric and a vacuum, the \(\vec{D}\) field obeys specific boundary conditions. If there is no free surface charge at the interface, the normal component of \(\vec{D}\) is continuous. That is, \(D_{1n} = D_{2n}\). If there is a free surface charge density \(\sigma_f\) at the interface, the discontinuity in the normal component is equal to that free charge density: \(D_{2n} - D_{1n} = \sigma_f\). These conditions are essential for solving problems involving multiple dielectric layers.