Supercapacitor

What a supercapacitor actually is

A supercapacitor — equally called an ultracapacitor or an electric double-layer capacitor (EDLC) — is a device that stores electrical energy by physically separating charge, not by running a chemical reaction. That single distinction is the root of everything interesting about it. A battery holds energy in chemical bonds; a supercapacitor holds it in the geometry of ions sitting close to a charged surface. Nothing is made and nothing is broken. Ions simply move toward, and pack against, the surface of two electrodes — and unpacking them is just as fast.

Inside the device are two porous carbon electrodes soaked in a liquid electrolyte, separated by a thin ion-permeable membrane. Apply a voltage and the electrolyte's ions sort themselves out: cations migrate to the negative electrode, anions to the positive one. At each electrode surface they form an electric double layer — a sheet of electronic charge in the carbon facing a counter-sheet of ionic charge in the electrolyte, separated by a gap of only a fraction of a nanometre. The supercapacitor is, in effect, two capacitors in series, one at each electrode, sharing the same electrolyte as the path between them.

That nanometre gap is the whole trick. A capacitor's stored charge is proportional to its plate area and inversely proportional to the distance between the plates. Make the plates microscopically porous so their internal surface is gigantic, and make the dielectric gap as thin as physics allows, and you get capacitance values — thousands of farads — that would be absurd for any conventional capacitor.

The electric double layer — Helmholtz to Gouy-Chapman-Stern

The structure of charge at a charged surface in an electrolyte was first modelled by Hermann von Helmholtz in 1853 as a simple parallel-plate capacitor: a rigid sheet of counter-ions held one ionic radius away from the electrode. That compact layer — now called the Helmholtz layer or Stern layer — is where most of a supercapacitor's charge lives, and it is the reason the effective plate separation d is sub-nanometre.

The full picture is the Gouy-Chapman-Stern model, which splits the electrolyte side into two regions: a compact Stern layer of ions essentially touching the surface, and a diffuse layer where thermal motion smears the remaining counter-ions out over a few nanometres. The total double-layer capacitance is the Stern and diffuse capacitances in series:

1 / C_dl = 1 / C_Stern + 1 / C_diffuse

For a typical activated-carbon / organic-electrolyte interface the specific double-layer capacitance is about 5 to 20 microfarads per square centimetre of real surface — which sounds tiny until you multiply it by thousands of square metres per gram.

Why the capacitance is so enormous

The governing relation is the parallel-plate capacitance with a relative permittivity for the electrolyte:

C = ε₀ · ε_r · A / d

ε₀  = 8.854 × 10⁻¹² F/m   (permittivity of free space)
ε_r ≈ 6–40                 (electrolyte near a charged wall, lower than bulk water's 80)
A   = total electrode surface area
d   ≈ 0.3–1 nm             (Helmholtz layer thickness)

Now plug in the supercapacitor numbers. A 1-gram activated-carbon electrode can present A ≈ 2,000 m² of accessible surface — about half a football field rolled up inside a grain of charcoal. With d ≈ 0.5 nm and ε_r ≈ 10:

C_single_electrode = ε₀ · ε_r · A / d
                   = (8.854e-12 · 10 · 2000) / (0.5e-9)
                   ≈ 350 F per gram   (idealised — ignores pore wetting,
                                        tortuosity, and inaccessible micropores)

Real cells reach a healthy fraction of that ideal because not every pore is small enough to wet, large enough to admit a solvated ion, or short enough to fill quickly — but even at partial pore utilisation you land at the hundred-farads-per-gram regime commercial cells actually operate in (100 to 350 F/g of carbon is typical). Multiply that by tens or hundreds of grams of carbon in a cell and you reach the thousands of farads on the datasheet. Note too that the cell capacitance is the two electrode capacitances in series, so a symmetric cell delivers half the single-electrode value:

1 / C_cell = 1 / C_pos + 1 / C_neg
For C_pos = C_neg = C:  C_cell = C / 2

Energy, power, and why the trade-off looks the way it does

Stored energy and peak power follow directly from the capacitor equations:

Energy:        E = ½ · C · V²
Peak power:    P_max = V² / (4 · ESR)
RC time:       τ = ESR · C

Two facts fall out of these. First, energy scales with voltage squared, which is why the low single-cell voltage (2.7 V) hurts so much and why so much engineering effort goes into raising it. A 3,000 F, 2.7 V Maxwell-style cell stores:

E = ½ · 3000 · 2.7²  =  10,935 J  ≈  3.04 Wh   (mass ≈ 510 g → ~6 Wh/kg)

Second, peak power is gated by the equivalent series resistance (ESR), which is tiny — often under 0.3 milliohm for a large cell. A 2.7 V cell with 0.29 mΩ ESR can in principle dump:

P_max = 2.7² / (4 · 0.00029)  ≈  6,300 W   from a single ~0.5 kg cell
                              → power density in the 10–14 kW/kg range

That is one to two orders of magnitude above a lithium-ion cell's power density, and it can be delivered in milliseconds because there is no reaction-rate bottleneck — only ions sliding into the double layer. The matching τ = ESR·C of roughly one second is what sets the realistic full-discharge timescale.

Supercapacitor vs lithium-ion vs film capacitor

Putting a supercapacitor next to the device it is most often confused with (a battery) and the device it descends from (a conventional capacitor) makes the niche obvious:

Property	Supercapacitor (EDLC)	Lithium-ion cell	Film / ceramic capacitor
Storage mechanism	Electrostatic double layer	Chemical intercalation	Dielectric polarisation
Specific energy	5–10 Wh/kg	150–250 Wh/kg	0.01–0.05 Wh/kg
Specific power	5,000–15,000 W/kg	200–1,000 W/kg	>100,000 W/kg
Charge time	1–60 s	30 min–4 h	microseconds
Cycle life	500k–1,000k	1,000–3,000	>1,000,000
Cell voltage	2.5–2.8 V	3.6–3.7 V	up to 1,000 V+
Round-trip efficiency	95–98 %	90–95 %	>99 %
Temperature range	−40 to +65 °C	0 to +45 °C ideal	−55 to +125 °C
Self-discharge	High (days–weeks)	Low (months)	Very low

The supercapacitor sits in the gap between the two: a thousand times more energy than a film capacitor, a thousand times more power and cycles than a battery. It is the bridge component you reach for when neither extreme fits.

How a commercial cell is built

A standard wound EDLC is built much like an electrolytic capacitor. Two strips of aluminium foil are coated with a slurry of activated carbon, conductive carbon black, and a binder (often PTFE or a fluoropolymer), then dried and calendered to a film about 100 micrometres thick. The two coated foils are wound around a separator — a thin, porous paper or polymer membrane that blocks electronic contact but passes ions freely. The jelly-roll is dropped into an aluminium can, vacuum-impregnated with electrolyte, and laser-welded shut.

• Electrode. Activated carbon dominates because it is cheap, conductive enough, chemically inert, and tunable in pore size. Newer electrodes use carbide-derived carbon, graphene, carbon nanotubes, or carbon aerogels to control pore geometry precisely — the goal is pores just barely larger than a desolvated ion, which paradoxically increases capacitance.
• Electrolyte. Organic electrolytes (tetraethylammonium tetrafluoroborate in acetonitrile or propylene carbonate) give the 2.7 V window and dominate commercial cells. Aqueous electrolytes (KOH, H₂SO₄) are cheaper, safer, and lower-resistance but cap out near 1 V. Ionic liquids push the window to 3.5 V+ but are viscous and expensive.
• Separator. Cellulose or polypropylene, 20–40 µm thick, providing the ion path while preventing a short.
• Current collector. Etched aluminium foil, sometimes carbon-coated to lower contact resistance — a dominant contributor to ESR.

Variants — EDLC, pseudocapacitor, hybrid

• Symmetric EDLC. Both electrodes are identical activated carbon; charge is stored purely electrostatically. Longest life, highest power, lowest energy. The workhorse of the industry.
• Pseudocapacitor. One or both electrodes use a material that undergoes a fast, reversible, surface-confined redox reaction — RuO₂ (the classic, but expensive), MnO₂, or conducting polymers like polypyrrole. The faradaic charge adds 2–10× the double-layer charge per area, raising energy density, but cycle life and power drop because a chemical step is now in the loop.
• Hybrid / lithium-ion capacitor (LIC). One battery-type electrode (graphite or lithium-titanate, pre-doped with lithium) paired with one carbon EDLC electrode. This raises cell voltage to 3.8 V and roughly triples energy density (to ~15–25 Wh/kg) while keeping much of the power and cycle advantage. The fastest-growing segment, used in regenerative and grid applications.
• Micro-supercapacitor. Thin-film, interdigitated electrodes patterned directly onto a chip or flexible substrate for on-board energy buffering in sensors and wearables.

Where supercapacitors earn their place

• Regenerative braking and transit. Trams, buses, and rubber-tyred gantry cranes capture kilojoules during seconds of braking and release them on acceleration. Shanghai's "capabus" recharges in about 30 seconds at each stop; Mazda's i-ELOOP captures braking energy into a 25 V capacitor bank to run the car's electrics.
• Engine start-stop and cold cranking. A capacitor delivers the high cranking current a battery struggles to provide at −30 °C, and absorbs the regenerative pulse, extending lead-acid battery life in start-stop vehicles.
• Grid and renewables. Frequency regulation, voltage ride-through, and the pitch-control backup that feathers wind-turbine blades during a grid fault — all need bursts of power for seconds, the supercapacitor's sweet spot.
• Power buffering and backup. Holding up memory and real-time clocks for days, smoothing the power draw of cellular radios and camera flashes, and bridging the milliseconds before a UPS or generator spins up.
• Motorsport KERS. Kinetic energy recovery systems briefly stored braking energy in supercapacitor banks for an acceleration boost before battery hybrids took over.

Failure modes and trade-offs

• Voltage overrun. Exceed the electrolyte's stability window and the solvent decomposes — the cell gasses, the can swells, ESR climbs, and capacitance collapses. This is why series stacks must be balanced.
• Cell imbalance in a stack. Real cells differ slightly in capacitance and leakage. In a series string the weakest cell sees the highest voltage and ages fastest, dragging down the whole bank. Passive balancing (bleed resistors) or active balancing (charge shuttling) is mandatory above a few cells.
• High self-discharge. A charged supercapacitor can lose a meaningful fraction of its charge in days through leakage and charge redistribution into deep pores. They are poor for long-term storage; they are buffers, not reservoirs.
• Voltage sag under load. Unlike a battery's flat discharge plateau, a capacitor's voltage falls linearly as it discharges (V = Q/C). Drawing usable energy down to half voltage already empties three-quarters of the stored energy, so the load electronics must tolerate a wide input range — usually via a DC-DC converter.
• ESR rise with age and temperature. Electrolyte dry-out and binder degradation slowly increase ESR, cutting deliverable power. End-of-life is often defined as a 100 % rise in ESR or a 20 % drop in capacitance.
• Energy-density ceiling. The fundamental limitation: even the best hybrids reach only ~25 Wh/kg. For applications that need sustained runtime rather than power bursts, a battery wins decisively, and no amount of better carbon closes that gap.

Common pitfalls when designing with supercapacitors

• Sizing on capacitance alone. Usable energy depends on the voltage window you can actually exploit. Pick the cutoff voltage and DC-DC range first, then size C and the series count.
• Ignoring balancing. A 48 V bank of 18 cells without balancing will kill its weakest cell within months. Budget for the balancing circuit from the start.
• Treating ESR as constant. ESR rises at low temperature and with age; the power available at −20 °C and end-of-life can be half the datasheet number. Design to worst case.
• Underestimating leakage and inrush. A discharged supercapacitor looks like a dead short at switch-on. Without inrush limiting it will trip protection or weld contacts. And leakage current means a charged bank needs periodic top-up.
• Expecting battery-like runtime. If the requirement is hours of energy, a supercapacitor is the wrong part. Match the device to the timescale: seconds and bursts, not hours.