Capacitors & Dielectrics · Guide · glidepathscience.org
GuideCapacitors & Dielectrics
Contents
This Guide explains the ideas and equations behind every Walkthrough step, Concept Challenge question and Predict round, in the order a textbook takes them. Part 1 prepares the Intro track and Part 2 the Advanced track; each takes about a quarter of an hour to read. Each section ends with a worked example that is not one of the sim’s own problems.
Part 1 · The essentials
Math you’ll need: algebra, powers of ten and unit prefixes (µ, n, p); one optional box uses an integral.
Two conductors and a field
A capacitor is any pair of conductors separated by an insulator. Connect it to a battery and charge flows: electrons leave one conductor and pile up on the other, until the voltage between them matches the battery’s. The conductors are now charged oppositely, and an electric field fills the space between them. Disconnect the battery and the charge stays where it is, ready to flow back through anything that connects the two conductors.
That simple object is everywhere. A camera flash dumps the energy of a capacitor into a lamp in a millisecond; a power supply smooths its output with capacitors; every phone holds hundreds of tiny ones; a touch screen and a condenser microphone sense their surroundings through a changing capacitance. This Guide follows the parallel-plate capacitor, two flat metal plates facing each other across a gap, because its field is uniform and every quantity can be worked out with a line or two of algebra. Then it fills the gap with an insulator and connects capacitors into networks.
Separated charge and Q = CV
When a battery charges a capacitor, it does not create charge: it moves charge from one plate to the other. One plate ends up with $+Q$ and the other with $−Q$, so the capacitor as a whole stays neutral. By convention “the charge on the capacitor” means $Q$, the charge on its positive plate. What a capacitor stores is separated charge, and with it energy.
How much charge moves depends on the voltage. Double the battery’s voltage and twice as much charge flows, because the field between the plates, and so the voltage across them, is proportional to the charge on them. The ratio is fixed for a given capacitor, and it is called the capacitance:
$C = Q/V$
Here $Q$ is the charge on the positive plate in coulombs and $V$ the potential difference between the plates in volts. Capacitance is measured in farads (F): one farad is one coulomb per volt. A farad is a great deal of capacitance, so most capacitors are rated in microfarads (µF, 10⁻⁶ F), nanofarads (nF, 10⁻⁹ F) or picofarads (pF, 10⁻¹² F). Rearranged, the definition gives the charge directly: $Q = CV$.
The capacitance itself depends on the shape and size of the conductors, as the next section shows.
Capacitance is geometry
For two flat plates of area $A$ a distance $d$ apart, much wider than the gap, the charge spreads evenly over the inner faces with density $σ = Q/A$ and makes a uniform field $E = σ/ε_0$ between them, where $ε_0$ = 8.854 × 10⁻¹² F/m is the permittivity of free space. The voltage across a uniform field is $V = Ed$. Putting these together, $V = Qd/(ε_0A)$, so
$C = ε_0A/d$
Each part makes sense. A larger area spreads the same charge more thinly, so the field and the voltage are smaller: more charge fits per volt. A wider gap means the same field acts over a longer distance, so the voltage is larger: less charge fits per volt. Notice what is not in the formula: neither $Q$ nor $V$. The definition $C = Q/V$ might suggest that more charge means more capacitance, but more charge always comes with proportionally more voltage. Capacitance is set by geometry and by what fills the gap, nothing else.
Real capacitors reach microfarads by shrinking the gap, folding a huge area into a small volume, and filling the gap with a dielectric; the Teardown shows four ways. First, though: what does a capacitor hold?
Energy stored, and where it lives
Charging a capacitor takes work. The first bit of charge crosses with no voltage to fight; the last crosses against the full voltage $V$. Because the voltage grows in proportion to the charge, the average charge moves against $V/2$, and the work done, now stored, is
$U = \tfrac12QV = \tfrac12CV^2 = Q^2/2C$
The three forms are the same thing, rewritten with $Q = CV$. Use whichever fits what you know. The factor ½ is not optional: $QV$ would be the work if the whole charge crossed the final voltage, which it never does.
Where is this energy? Not in the charges themselves but in the field between the plates. For the parallel plates, $\tfrac12CV^2 = \tfrac12(ε_0A/d)(Ed)^2 = \tfrac12ε_0E^2 × Ad$, and $Ad$ is the volume between the plates. So the field carries an energy density, in joules per cubic meter,
$u_E = \tfrac12ε_0E^2$
This is true of every electric field, not just a capacitor’s.
Changing a charged capacitor changes this energy, and how depends on whether the battery is still attached.
What is held fixed
There are two ways to change a charged capacitor, and they behave oppositely. If the battery stays connected, it holds the voltage fixed: any change in $C$ makes charge flow to or from the battery until $Q = CV$ again. If the battery is disconnected first, the capacitor is isolated and its charge is fixed, since there is nowhere for charge to go; the voltage then follows $V = Q/C$.
Pull the plates apart, doubling the gap, so $C$ halves. Isolated: $Q$ stays, $V$ doubles, the field $E = V/d$ stays the same (the same charge density makes the same field), and $U = Q^2/2C$ doubles. The extra energy is the work your hand did against the plates’ attraction. Connected: $V$ stays, $Q = CV$ halves (the excess charge is pushed back into the battery), $E$ halves, and $U = \tfrac12CV^2$ halves. You still did work against the attraction.
The same two cases decide what happens when the gap is filled with something other than air.
Dielectrics
Fill the gap with an insulator, a dielectric, and the capacitance rises by a factor $κ$ (kappa), the material’s dielectric constant:
$C = κC_0$
where $C_0$ is the capacitance with nothing (vacuum, or for practical purposes air) in the gap. For paper $κ$ is about 3.7, for glass about 5.6, for water about 80. With the battery connected, the extra capacitance draws more charge from it. Isolated, the charge is fixed and the voltage falls: $V = V_0/κ$, and so does the field.
Why? A dielectric is not a conductor: its electrons are bound to their molecules and cannot leave them. But in a field each molecule’s charges shift a little, the electrons against the field and the nuclei with it, and the molecule becomes a small dipole. Inside the material the shifted charges cancel each other, but at the faces they do not: a thin layer of negative bound charge is left on the face next to the positive plate, and positive bound charge on the other face. These bound charges make a field of their own, opposite to the plates’, so the field inside is smaller, by exactly the factor $κ$. A smaller field across the same gap means a smaller voltage for the same free charge: more capacitance.
An isolated capacitor’s energy falls as the dielectric slides in, $U = Q^2/2C$, because the field pulls the dielectric in and does work on it.
Breakdown
No insulator holds an unlimited field. Above a field called its dielectric strength, $E_c$, a material breaks down: in air, stray electrons gain enough energy between collisions to knock more electrons loose, and a spark crosses. Air’s strength is about 3.0 MV/m; paper’s about 16 MV/m; good plastic films hundreds of megavolts per meter. Since the field in a full gap is $V/d$, the highest voltage a gap can hold is
$V_{max} = E_cd$
This is why real capacitors are filled. A dielectric raises the capacitance by $κ$ and the voltage limit by the ratio of strengths, and the stored energy, $\tfrac12CV^2$, rises by both together. A capacitor’s voltage rating is set well below its breakdown voltage, by a safety factor.
Series and parallel
Circuits combine capacitors in two basic ways. In parallel, each capacitor is connected across the same two points, so each has the same voltage $V$. Their charges add: $Q = C_1V + C_2V + …$, so the combination acts as one capacitor of
$C_P = C_1 + C_2 + …$
It is as if the plates were placed side by side to make one larger plate. In series, the capacitors are connected end to end. The plates between neighbors form an island that started neutral and is connected to nothing else, so whatever charge gathers on one side of it leaves an equal and opposite charge on the other: every capacitor in the string carries the same charge $Q$. Their voltages add, $V = Q/C_1 + Q/C_2 + …$, so
$1/C_S = 1/C_1 + 1/C_2 + …$
A series string has less capacitance than its smallest member, as if the gaps were stacked into one wider gap. These rules are the reverse of the rules for resistors, a common slip. And since each part in series has the same $Q$, the voltage $V = Q/C$ is largest across the smallest capacitor: that is the one to check against its rating.
Part 2 · Going further
Math you’ll need: calculus (integrals and derivatives) and Gauss’s law; logarithms; solving two linear equations together.
Capacitance from Gauss’s law, and the edges
Every capacitance follows the same four steps: assume a charge $Q$ on one conductor (and $−Q$ on the other), find the field from Gauss’s law, integrate the field from one conductor to the other to find $V$, and divide. For a coaxial cable with charge $λ$ per meter on an inner conductor of radius $a$, inside an outer one of radius $b$, a cylinder of radius $r$ encloses $λ$ per meter, so $E = λ/(2πκε_0r)$. Integrating from $a$ to $b$ gives $V = (λ/2πκε_0)\ln(b/a)$, and
$C′ = 2πκε_0/\ln(b/a)$
per meter of cable. The same steps give two concentric spheres, $C = 4πε_0ab/(b − a)$, and an isolated sphere ($b → ∞$), $C = 4πε_0R$.
The parallel-plate formula assumed plates much wider than their gap, with no field outside. Real plates have edges, and the field bulges out around them, carrying extra charge near the ends and even on the plates’ outer faces. For long plates seen end-on (a 2-D problem), the field module solves this exactly: the extra capacitance is about 36 % at $L/d$ = 4, 20 % at 8 and 9 % at 20. H. B. Palmer’s 1937 formula, $C′/ε_0 = L/d + (1 + \ln(2πL/d))/π$, comes within a few percent of the extra once $L/d$ is 8 or more.
Forces and the energy books
The plates of a charged capacitor attract. Each plate feels only the other plate’s field, $σ/2ε_0$, half the field between them, so the force is $F = Q(E/2)$, or
$F = \tfrac12ε_0E^2A$
The pressure on the plates, $\tfrac12ε_0E^2$, equals the energy density. The same result comes from virtual work: at constant charge, $F = −dU/dd$, the rate at which the stored energy changes as the plate moves. Virtual work is often the easier route. A dielectric slab is pulled into a charged gap with $F = \tfrac12V^2\,dC/dx$, and a liquid dielectric between vertical plates rises until the pull, $\tfrac12V^2\,dC/dh$, balances the weight of the raised column: $h = (κ − 1)ε_0E^2/(2ρg)$. Only the excess over air, $κ − 1$, pulls; and the field is limited by the air above the liquid, which sparks first.
With the battery connected the books need care. Pull the plates apart slowly: $C$ falls, so $U = \tfrac12CV^2$ falls by $\tfrac12V^2(C_1 − C_2)$, while the charge pushed back through the battery, $V(C_1 − C_2)$ coulombs, delivers $V^2(C_1 − C_2)$ of energy into it, twice what the capacitor lost. The difference is your work:
$W = \tfrac12V^2(C_1 − C_2)$
Capacitive sensing
Anything that changes a capacitance can be measured through it. At constant charge, $V = Qd/(ε_0A)$ follows the gap: a condenser microphone biases its diaphragm through a resistor so large that $RC$ is far longer than a sound’s period, so the charge cannot follow and the output is $ΔV = (V_0/d)Δx$. Below the frequency $1/(2πRC)$ the charge does follow, and the microphone stops responding. At constant voltage, the charge drawn measures $C$ instead: touch screens sense the capacitance a fingertip adds, and fuel gauges the liquid between two electrodes.
Bound charge, D and layers
The bound charge on a dielectric’s faces is a fixed fraction of the free charge on the plates:
$σ_b = (1 − 1/κ)σ_f$
It approaches the free charge as $κ$ grows (a conductor is the limit) but never exceeds it. Two fields describe this. The displacement $D$ depends on free charge only: $D = σ_f$ between the plates, whatever fills them. The polarization $P$, the dipole moment per volume, equals $σ_b$. The field is what both make together, $E = (σ_f − σ_b)/ε_0 = D/(κε_0)$.
Gaps filled partly are capacitors in combination. Layers stacked across the gap carry the same $D$, so they are in series, each adding its thickness over its $κ$: $C = ε_0A/(d − t + t/κ)$ for a slab of thickness $t$ on one plate. Regions side by side carry the same $E$ (the same $V$ over the same $d$), so they are in parallel. In a stack the field is $κ$ times stronger in the air than in the slab, and air is the weaker material: an air layer or a bubble breaks down first, and adding a thin dielectric can lower the voltage a gap will hold. This is why voids in high-voltage insulation are dangerous.
κ from molecules
In a field $E$, a molecule’s electron cloud shifts and the molecule acquires a dipole moment $p = αE$, where $α$ is its polarizability. Tables often give the polarizability volume $α′ = α/(4πε_0)$, in ų (10⁻³⁰ m³). For a dilute gas of $N$ molecules per cubic meter the polarization is $P = NαE$, and
$κ − 1 = Nα/ε_0 = 4πNα′$
with $N = p/k_BT$ from the ideal gas law. Air comes out at $κ$ = 1.00054 at 20 °C. In a dense liquid each molecule also feels its neighbors’ dipoles; the Clausius–Mossotti relation, $(κ − 1)/(κ + 2) = (4π/3)Nα′$, accounts for that. For water it gives $κ$ = 1.80, close to the square of water’s refractive index, $n^2$ = 1.78. Yet water’s $κ$ is 80. The difference is orientation: each water molecule has a permanent dipole, and the field turns the dipoles partly into line against thermal jostling. That effect falls as the water warms, and vanishes at optical frequencies, where the dipoles cannot turn fast enough: $n^2$ is all that remains.
Real capacitors
Practical capacitors reach large capacitances with thin dielectrics, large areas and high $κ$. A film capacitor rolls two metallized plastic strips together, so both faces of each strip work. A multilayer ceramic chip stacks hundreds of thin layers of a high-$κ$ ceramic in parallel. An aluminum electrolytic grows an oxide tens of nanometers thick on etched foil, multiplying its area. A double-layer capacitor stores charge as ions a nanometer from a carbon sponge’s enormous surface.
Two material properties set the limits. No insulator is perfect: with resistivity $ρ$, the leakage resistance $R = ρd/A$ and $C = κε_0A/d$ give a self-discharge time $τ = RC = κε_0ρ$, the same for any size. And the most energy a dielectric can hold per volume, at its breakdown field, is $u_{max} = \tfrac12κε_0E_c^2$: the strength enters squared, so a strong film beats a high-$κ$ ceramic.
Networks by nodes
Some networks are neither series nor parallel, a bridge for one. They still follow from two rules: each capacitor obeys $Q = CV$ across it, and each island (a set of plates joined to each other but to nothing else) keeps the net charge it started with, zero for capacitors that start uncharged. Write the island’s charge, $Σ C_i(V_{island} − V_{neighbor}) = 0$, for each island, solve the equations together for the islands’ potentials, and every charge follows. The equivalent capacitance is the charge drawn from the battery divided by its voltage.
Reference
Common mistakes
Part 1
- Saying a capacitor holds net charge: it holds $+Q$ and $−Q$, separated.
- Thinking $C$ depends on $Q$ or $V$: it depends on the geometry and the fill.
- Writing $U = QV$: the charge crossed an average of $V/2$, so $U = \tfrac12QV$.
- Holding the voltage fixed for an isolated capacitor: with the battery gone, $Q$ is fixed.
- Thinking a battery holds the charge fixed: it holds the voltage, and charge flows.
- Treating a dielectric as a weak conductor: its charges are bound, and only shift.
- Swapping the series and parallel rules with the resistor rules.
- Giving the larger capacitor the larger voltage in series: same $Q$, so the smaller one takes more.
Part 2
- Using $QE$ with the whole field for the force between plates: each plate feels $E/2$.
- Using $κ$ where $κ − 1$ belongs: air already fills the gap with $κ$ = 1.
- Thinking any added dielectric raises the breakdown voltage: an air layer in series fails first.
The equation sheet
| Part 1 | |
|---|---|
| capacitance | $Q = CV$ |
| parallel plates | $C = κε_0A/d$ |
| a uniform field | $E = V/d$ |
| energy | $U = \tfrac12QV = \tfrac12CV^2 = Q^2/2C$ |
| energy density | $u = \tfrac12ε_0E^2$ |
| a dielectric, isolated | $V = V_0/κ$ (isolated) |
| breakdown | $V_{max} = E_cd$ |
| parallel | $C_P = C_1 + C_2 + …$ |
| series | $1/C_S = 1/C_1 + 1/C_2 + …$ |
| Part 2 | |
|---|---|
| coax | $C′ = 2πκε_0/\ln(b/a)$ |
| the force between plates | $F = \tfrac12ε_0E^2A$ |
| your work, battery on | $W = \tfrac12V^2(C_1 − C_2)$ at constant $V$ |
| bound charge | $σ_b = (1 − 1/κ)σ_f$ |
| a stacked layer | $C = ε_0A/(d − t + t/κ)$ |
| a liquid’s rise | $h = (κ − 1)ε_0E^2/(2ρg)$ |
| a dilute gas | $κ − 1 = Nα/ε_0 = 4πNα′$ with $N = p/k_BT$ |
| a microphone | $ΔV = V_0Δx/d$ |
Symbols: $Q$ the charge on the positive plate; $V$ the voltage across; $C$ the capacitance ($C_0$ empty); $A$ the plates’ area; $d$ the gap; $ε_0$ = 8.854 × 10⁻¹² F/m; $κ$ the dielectric constant; $E$ the field and $E_c$ a material’s strength; $U$ and $u$ the energy and energy density; $σ_f$, $σ_b$ the free and bound charge densities; $t$ a layer’s thickness; $a$, $b$ radii; $ρ$ a density (or resistivity); $g$ = 9.81 m/s²; $N$ molecules per volume, $α$ the polarizability and $α′$ its volume; $k_B$ Boltzmann’s constant; $T$ the temperature in kelvin; $V_0$ a bias and $Δx$ a displacement.
Sources and further reading
- OpenStax, University Physics Volume 2, chapter 8, “Capacitance,” CC BY-NC-SA 4.0 (openstax.org): the Intro track follows its order and notation.
- E. M. Purcell and D. J. Morin, Electricity and Magnetism, 3rd ed. (Cambridge, 2013), and D. J. Griffiths, Introduction to Electrodynamics, 5th ed. (Cambridge, 2023), chapter 4, for dielectrics, bound charge and forces by virtual work.
- H. B. Palmer, “The capacitance of a parallel-plate capacitor by the Schwartz–Christoffel transformation,” Trans. AIEE 56, 363 (1937).
- The CRC Handbook of Chemistry and Physics for typical dielectric constants, strengths and polarizabilities.
- This sim’s Field Manual (H) for the model and its limits; the RC Circuit Lab for charging in time; Parallel Plates & Uniform Fields for the uniform field itself.
Capacitors & Dielectrics v1.1 · © 2026 Dr. David Meier · CC BY-NC-SA 4.0 · glidepathscience.org