Skip to content

Unit 10: Conductors and Capacitors

Dielectrics

Explain how an insulator between capacitor plates polarizes, weakens the field and raises capacitance by κ, and predict changes with or without a battery.

Topic 10.4 · 9 min read

Fill the gap of a charged, isolated capacitor with a sheet of glass and a voltmeter across the plates drops: no charge left the plates, yet the potential difference fell. The glass is a dielectric, an insulator placed in an electric field, and it weakens the field between the plates without cancelling it. That is how real capacitors pack microfarads into a component the size of a grain of rice.

How an insulator polarizes

In a conductor (topic 10.1), electrons roam the whole object and pile up on its surface until the field inside is exactly zero. In an insulator, every electron stays bound to its own atom or molecule. An external field can only shift charge a tiny distance within each molecule, in one of two ways:

  • Induced dipoles. In a nonpolar molecule (like the long chains in Teflon or polystyrene), the field pushes the electron cloud one way and the nuclei the other, stretching the molecule into a small dipole aligned with the field.
  • Aligned permanent dipoles. A polar molecule, such as water, is already a dipole. The field pulls its positive end one way and its negative end the other, which twists the molecule partly into line. Thermal jostling keeps the alignment from being complete.

Either way the material becomes polarized: its molecules point their negative ends toward the positive plate and their positive ends toward the negative plate. Deep inside, each molecule's positive end sits next to its neighbor's negative end, and the charges cancel on average. At the two faces they do not: the face next to the positive plate is left with a layer of negative charge, and the face next to the negative plate a layer of positive charge. These bound charges can't leave their molecules, unlike the free charges on the plates, which can flow through wires. The dielectric as a whole stays neutral.

A parallel-plate capacitor with a dielectric slab between the plates. The top plate carries free charge plus Q and the bottom plate free charge minus Q. Inside the slab, molecules drawn as small ovals are all aligned with their negative ends up and positive ends down. A row of bound negative charge lines the slab's top face, next to the positive plate, and a row of bound positive charge lines its bottom face. An arrow labeled E zero points down, from the positive plate toward the negative plate; a shorter arrow labeled E induced points up, opposite to it.

The dielectric's field opposes the applied field

The layer of bound charge acts like a second, weaker pair of plates inside the first, with the signs reversed. Its field, E⃗ind\vec{E}_{\text{ind}}, points from the bound positive face to the bound negative face, which is opposite to the applied field, the field E⃗0\vec{E}_0 of the free charge on the plates. The net field in the dielectric is smaller than E⃗0\vec{E}_0 but in the same direction, because the bound charge is always less than the free charge (the molecules only partly cancel the field that is polarizing them).

The dielectric constant κ\kappa (kappa) measures how much a material weakens the field. For a dielectric filling the gap of an isolated capacitor,

E=E0κ.E = \frac{E_0}{\kappa}.

The free surface charge density σ\sigma on the plates and the bound density σb\sigma_b on the dielectric's faces add like two pairs of sheets, so E=σ−σbε0E = \frac{\sigma - \sigma_b}{\varepsilon_0}. Setting this equal to σκε0\frac{\sigma}{\kappa\varepsilon_0} gives σb=σ(1−1κ)\sigma_b = \sigma\left(1 - \frac{1}{\kappa}\right): with κ=4\kappa = 4, the bound charge is three quarters of the free charge and the field is a quarter as strong.

Some values of κ\kappa at room temperature:

Materialκ\kappa
Vacuum1 (exactly)
Dry air1.0006
Teflon2.1
Paper3.7
Water80

Every material has κ>1\kappa > 1, because polarization can only oppose the applied field. Air is so close to vacuum that the exam treats air-filled capacitors as having κ=1\kappa = 1. Water's large value comes from its strongly polar molecules, which turn readily into line.

The dielectric constant also defines the material's permittivity, ε=κε0\varepsilon = \kappa\varepsilon_0. Inside a dielectric, every formula you derived with ε0\varepsilon_0 carries over with ε\varepsilon in its place.

What changes when a dielectric goes in

The capacitance follows directly from the weaker field. With the plates' charge QQ fixed, ΔV=Ed=E0dκ=ΔV0κ\Delta V = Ed = \frac{E_0 d}{\kappa} = \frac{\Delta V_0}{\kappa}, so C=QΔV=κC0C = \frac{Q}{\Delta V} = \kappa C_0. The capacitance is a property of the device (topic 10.3), so it rises by κ\kappa in every case. What happens to the other quantities depends on whether the capacitor is isolated or stays connected to a battery:

Filled with κ\kappaCCQQΔV\Delta VEEUU
Isolated (battery removed first)×κ\times\kappasame÷κ\div\kappa÷κ\div\kappa÷κ\div\kappa
Connected to battery×κ\times\kappa×κ\times\kappasamesame×κ\times\kappa

Isolated: the charge has nowhere to go, so the dielectric's field weakens the net field and the potential difference falls by κ\kappa. The energy Q22C\frac{Q^2}{2C} falls by κ\kappa as well. The missing energy went into the dielectric: the plates' field pulls a dielectric slab into the gap (the bound charges near the edge are attracted toward the plates of opposite sign), so you must hold it back as it goes in, and the field does positive work on it.

Connected: the battery fixes ΔV\Delta V, so the field ΔVd\frac{\Delta V}{d} between the plates can't change. As the dielectric starts to weaken the field, the battery pushes more charge onto the plates until the field is back to ΔVd\frac{\Delta V}{d}; the free charge ends at κQ0\kappa Q_0. The energy 12C(ΔV)2\frac{1}{2}C(\Delta V)^2 rises by κ\kappa, supplied by the battery.

The general statement is that inserting a dielectric may change each quantity. Decide first what is held fixed, as in topic 10.3's Example 2.

Worked examples

Worked exampleExample 1: glass into an isolated capacitor

An air-filled parallel-plate capacitor with C0=50 pFC_0 = 50\ \text{pF} and plate separation 2.0 mm2.0\ \text{mm} is charged to 200 V200\ \text{V} and disconnected from the battery. A glass slab with κ=4.0\kappa = 4.0 is slid in, filling the gap. Find the new capacitance, charge, potential difference, field and stored energy, and the bound charge on the face of the glass.

Show solution

Disconnected, so the plates' charge is fixed. Start there:

Q=C0 ΔV0=(50×10−12 F)(200 V)=1.0×10−8 C.Q = C_0\,\Delta V_0 = (50\times10^{-12}\ \text{F})(200\ \text{V}) = 1.0\times10^{-8}\ \text{C}.

The glass fills the gap, so C=κC0=200 pFC = \kappa C_0 = 200\ \text{pF}, and

ΔV=QC=1.0×10−8200×10−12=50 V.\Delta V = \frac{Q}{C} = \frac{1.0\times10^{-8}}{200\times10^{-12}} = 50\ \text{V}.

The field was E0=200 V2.0×10−3 m=1.0×105 V/mE_0 = \frac{200\ \text{V}}{2.0\times10^{-3}\ \text{m}} = 1.0\times10^{5}\ \text{V/m} and is now 50 V2.0×10−3 m=2.5×104 V/m\frac{50\ \text{V}}{2.0\times10^{-3}\ \text{m}} = 2.5\times10^{4}\ \text{V/m}, one quarter as large, still pointing from the positive plate to the negative plate.

Energy before: U0=12C0(ΔV0)2=12(50×10−12)(200)2=1.0×10−6 JU_0 = \frac{1}{2}C_0(\Delta V_0)^2 = \frac{1}{2}(50\times10^{-12})(200)^2 = 1.0\times10^{-6}\ \text{J}. After, using the form with the fixed quantity: U=Q22C=(1.0×10−8)22(200×10−12)=2.5×10−7 JU = \frac{Q^2}{2C} = \frac{(1.0\times10^{-8})^2}{2(200\times10^{-12})} = 2.5\times10^{-7}\ \text{J}. The 7.5×10−7 J7.5\times10^{-7}\ \text{J} difference is the work the field did pulling the glass in.

The bound charge on each face of the glass is Q(1−1κ)=(1.0×10−8)(0.75)=7.5×10−9 CQ\left(1 - \frac{1}{\kappa}\right) = (1.0\times10^{-8})(0.75) = 7.5\times10^{-9}\ \text{C}: negative on the face next to the positive plate.

Worked exampleExample 2: a slab that fills half the gap

A parallel-plate capacitor has plates of area 0.010 m20.010\ \text{m}^2 separated by 4.0 mm4.0\ \text{mm} and stays connected to a 90 V90\ \text{V} battery. A plastic slab with κ=3.0\kappa = 3.0, 2.0 mm2.0\ \text{mm} thick and the same area as the plates, is placed flat against one plate. Find the field in the air and in the plastic, the charge on the plates, and the new capacitance. Use ε0=8.85×10−12 C2/(N⋅m2)\varepsilon_0 = 8.85\times10^{-12}\ \text{C}^2/(\text{N}\cdot\text{m}^2).

Show solution

The formula C=κC0C = \kappa C_0 is for a filled gap, so go back to the field. The free charge density σ\sigma on the plates is the same all the way across, so the field in the air gap is Eair=σε0E_{\text{air}} = \frac{\sigma}{\varepsilon_0} and in the plastic it is reduced by κ\kappa: Eplastic=Eair3.0E_{\text{plastic}} = \frac{E_{\text{air}}}{3.0}. The battery fixes the total potential difference, which is the sum across the two layers:

90 V=Eair(2.0×10−3 m)+Eair3.0(2.0×10−3 m)⇒Eair=3.375×104 V/m.90\ \text{V} = E_{\text{air}}(2.0\times10^{-3}\ \text{m}) + \frac{E_{\text{air}}}{3.0}(2.0\times10^{-3}\ \text{m}) \quad\Rightarrow\quad E_{\text{air}} = 3.375\times10^{4}\ \text{V/m}.

Then Eplastic=1.125×104 V/mE_{\text{plastic}} = 1.125\times10^{4}\ \text{V/m}. Both point from the positive plate to the negative plate.

The charge comes from the air-gap field: Q=σA=ε0EairA=(8.85×10−12)(3.375×104)(0.010)=2.99×10−9 CQ = \sigma A = \varepsilon_0 E_{\text{air}} A = (8.85\times10^{-12})(3.375\times10^{4})(0.010) = 2.99\times10^{-9}\ \text{C}.

The capacitance is C=QΔV=2.99×10−990=3.32×10−11 FC = \frac{Q}{\Delta V} = \frac{2.99\times10^{-9}}{90} = 3.32\times10^{-11}\ \text{F}. Without the slab, C0=ε0Ad=2.21×10−11 FC_0 = \frac{\varepsilon_0 A}{d} = 2.21\times10^{-11}\ \text{F}, so the half-thickness slab raised CC by a factor of 1.51.5, not 3.03.0. A metal slab of the same thickness would raise it by a factor of 2, since it removes the field from its whole thickness (topic 10.1, Example 3); a dielectric with a very large κ\kappa behaves almost like that metal.

Worked exampleExample 3: measuring a dielectric constant (experimental design)

Students want to find the dielectric constant of a plastic that comes in sheets 0.50 mm0.50\ \text{mm} thick. They have two square metal plates of area 0.0400 m20.0400\ \text{m}^2, a stack of the sheets, and a capacitance meter.

(a) Describe a procedure that would let them determine κ\kappa from a graph. Say what they vary, what they measure and what they keep the same.

(b) Their data are below. What should they graph to get a straight line, and how is κ\kappa found from it?

Sheetsdd (mm)CC (nF)
10.502.30
21.001.12
31.500.76
42.000.56
52.500.46

(c) Use the data to calculate κ\kappa. Use ε0=8.85×10−12 C2/(N⋅m2)\varepsilon_0 = 8.85\times10^{-12}\ \text{C}^2/(\text{N}\cdot\text{m}^2).

(d) In one trial, thin air gaps were left between the sheets. Would that make the measured value of κ\kappa too high or too low? Explain.

Show solution

(a) Stack 1, 2, 3, 4 and 5 sheets between the plates, pressing the plates flat so the plastic fills the gap with no air. For each stack, record the plate separation (number of sheets times 0.50 mm0.50\ \text{mm}, or measured with calipers) and the capacitance from the meter. Keep the plates and the plastic the same throughout, and keep the plates square to each other so the area of overlap stays 0.0400 m20.0400\ \text{m}^2. Take several readings per stack and average them.

(b) The model is C=κε0AdC = \frac{\kappa\varepsilon_0 A}{d}, so CC is proportional to 1d\frac{1}{d}. Graph CC against 1d\frac{1}{d}: the points should fall on a line through the origin with slope κε0A\kappa\varepsilon_0 A, so κ=slopeε0A\kappa = \frac{\text{slope}}{\varepsilon_0 A}. A graph of CC against dd would be a curve, which is much harder to read a constant from.

Graph of capacitance in nanofarads against one over d in inverse meters. Five data points, at 1 over d equal to 400, 500, 667, 1000 and 2000 inverse meters with C equal to 0.46, 0.56, 0.76, 1.12 and 2.30 nanofarads, lie close to a straight best-fit line through the origin with slope about 1.15 times 10 to the minus 12 farad meters.

(c) The 1d\frac{1}{d} values are 20002000, 10001000, 667667, 500500 and 400 m−1400\ \text{m}^{-1}. Draw the best-fit line through the plotted points and read two well-separated points on the line (not data points), such as (400 m−1, 0.45 nF)(400\ \text{m}^{-1},\ 0.45\ \text{nF}) and (2000 m−1, 2.29 nF)(2000\ \text{m}^{-1},\ 2.29\ \text{nF}):

slope=(2.29−0.45)×10−9 F(2000−400) m−1=1.15×10−12 F⋅m.\text{slope} = \frac{(2.29 - 0.45)\times10^{-9}\ \text{F}}{(2000 - 400)\ \text{m}^{-1}} = 1.15\times10^{-12}\ \text{F}\cdot\text{m}.

Then

κ=1.15×10−12 F⋅m(8.85×10−12 C2/(N⋅m2))(0.0400 m2)=3.2.\kappa = \frac{1.15\times10^{-12}\ \text{F}\cdot\text{m}}{(8.85\times10^{-12}\ \text{C}^2/(\text{N}\cdot\text{m}^2))(0.0400\ \text{m}^2)} = 3.2.

(Units: ε0A\varepsilon_0 A is in F⋅\cdotm, so κ\kappa is dimensionless, as it must be.) Taking the slope from the line, rather than from two of the data points, averages out the scatter in the measurements.

(d) Too low. In an air gap the field is not reduced by the plastic, so, as in Example 2, a gap partly filled with plastic has less capacitance than one completely filled. The measured CC for that stack comes out smaller than κε0Ad\frac{\kappa\varepsilon_0 A}{d}, which pulls the slope, and the calculated κ\kappa, down.

Where students lose points

On the exam

Dielectric questions ask you to predict changes in CC, QQ, ΔV\Delta V, EE and UU when a dielectric is inserted or removed, to explain polarization and the direction of the induced field, and to design or analyze an experiment that measures κ\kappa (linearize the data, take the slope from a best-fit line, carry units). A calculator is allowed throughout.

Check yourself

  1. Question 1

    An air-filled parallel-plate capacitor is charged and disconnected from the battery. A slab with κ=2\kappa = 2 is then slid in, filling the gap. Which describes the result?

    1. A

      QQ doubles, and ΔV\Delta V is unchanged.

    2. B

      QQ is unchanged, and the field between the plates is halved.

    3. C

      QQ is unchanged, and the field between the plates is unchanged.

    4. D

      QQ is halved, and ΔV\Delta V is halved.

    Show answer

    B

    Disconnected, the free charge can't change. The polarized slab's bound charge sets up a field opposite the plates' field, so the net field drops to E0κ\frac{E_0}{\kappa} and ΔV\Delta V drops with it. Choice A describes the connected case, where the battery holds ΔV\Delta V fixed.

  2. Question 2

    A parallel-plate capacitor stays connected to a battery while a dielectric with κ=5\kappa = 5 is inserted to fill the gap. By what factors do the charge on the plates, the field between the plates and the stored energy change?

    Show answer

    The charge increases by a factor of 5, the field is unchanged, and the stored energy increases by a factor of 5.

    The battery fixes ΔV\Delta V, so E=ΔVdE = \frac{\Delta V}{d} can't change. CC becomes 5C05C_0, so Q=C ΔVQ = C\,\Delta V becomes 5Q05Q_0 and U=12C(ΔV)2U = \frac{1}{2}C(\Delta V)^2 becomes 5U05U_0; the battery supplies the extra charge and energy.

  3. Question 3

    A dielectric slab fills the gap between a positively charged upper plate and a negatively charged lower plate. Which statement about the slab is correct?

    1. A

      Its top face carries negative bound charge, and the field of its bound charge points up.

    2. B

      Its top face carries positive bound charge, and the field of its bound charge points down.

    3. C

      Its top face carries negative bound charge, and the field of its bound charge points down.

    4. D

      It carries no surface charge, because its charges cannot move.

    Show answer

    A

    The molecules turn or stretch so their negative ends face the positive plate, leaving negative bound charge on the top face and positive on the bottom. The field of that bound charge points from the bottom (positive) face to the top (negative) face, up, opposite the plates' downward field. Choice D is wrong because bound charges shift slightly within each molecule, and that shift leaves charge at the faces.

  4. Question 4

    A parallel-plate capacitor has plates of area 0.050 m20.050\ \text{m}^2 separated by 0.20 mm0.20\ \text{mm} of a material with κ=2.5\kappa = 2.5 that fills the gap. Find its capacitance. Use ε0=8.85×10−12 C2/(N⋅m2)\varepsilon_0 = 8.85\times10^{-12}\ \text{C}^2/(\text{N}\cdot\text{m}^2).

    Show answer

    C≈5.5×10−9 FC \approx 5.5\times10^{-9}\ \text{F} (about 5.5 nF5.5\ \text{nF}).

    C=κε0Ad=(2.5)(8.85×10−12)(0.050)2.0×10−4=5.5×10−9 FC = \frac{\kappa\varepsilon_0 A}{d} = \frac{(2.5)(8.85\times10^{-12})(0.050)}{2.0\times10^{-4}} = 5.5\times10^{-9}\ \text{F}. The material raises the capacitance of the same plates in air by a factor of 2.5.

Go deeper